City of RomeCity of Rome

O. Hänninen ¹ , P. Lipponen ¹ , R. Sorjamaa ¹ , M. Lamberti ² , C. Gariazzo ²

1

National Institute for Health and Welfare (THL)

E X P A H

Action 5.1 – Statistical i/o –analysis Action 5.2 – Modelling infiltration Action 5.3 – Exposure modelling

Time-activity, sources and infiltration

1

National Institute for Health and Welfare (THL)

2

INAIL

Department of Environmental Health

Talk outline

• Exposures determine health effects!

• What are the main sources of exposures? Role of traffic?

• Where are the main contacts with population time-activity?

• How do outdoor PAHs infiltrate indoors?

How to model PAH exposures?

• How do outdoor PAHs infiltrate indoors?

• Population spends substantial fraction of time indoors. Are people safe indoors? Are there indoor sources?

• Exposure modelling

Exposure and effect chain

Man Environment

Exposure Health

Concentrations Doses

Sources

Ambient

epidemiology

Ambient Outdoors Indoors Intake Uptake

Cumulative retained

dose

Exposure Health

Concentrations Doses

Emissions

Sources

Microenvironments

Analysis of

Exposure determinants 1

• What are the main sources of exposures in Rome?

=> Which microenvironments are critical for exposures?

• Traffic?

• Outdoors?

• Outdoors?

• Indoors: Homes, schools, offices?

Emission inventory

suggested three main sources

C it y o f R o m e

Source: EXPAH emission inventory (Radice et al., 2012)

C it y o f R o m e

(IP)

Σ4PAH 87%

Four PAH compounds

Measurements

• Four microenvironment types:

– homes (8+1/1) – schools (7)

– offices (3) – vehicles (3)

EXPAH Actions 3

– vehicles (3)

• indoors and outdoors (146

successful measurement pairs)

• 1-7 day sampling

with PM _2.5 inlets Microenvironments Sites

Home ^8+1/1

Office 3

School 7

Vehicles ³

Compounds

• 11 individual PAs

• 3 integrated groups (optional)

Groups Compounds

PAH compounds 1 BaA

2 BaP 3 CH 4 BPE 5 DBA 6 IP 7 PE 8 BeP 9 BbF 10 BjF 11 BkF Integrated peaks BbjF

BjkF BbjkF

Four PAH compounds in emission inventory EXPAH Actions 3

• 5 summed PAH groups

• 3 calculated ratios

• 3 other PM-bound pollutants

BbjkF

Summed groups ΣBbjF

ΣBbjkF ΣBjkF

PAHs (4 compounds) ΣPAH (all compounds)

Ratios BaP/ΣPAH

BaP/PAHs BaP/BeP Other pollutants PM2.5

EC

OC

25 30 35 40 45 50

P A H c o n ce n tr a ti o n ( n g m -3 )

ΣPAH indoor ΣPAH outdoor

Weekly PAH concentrations by season

Heating season

Initial observations:

• i/o ratios 0.5 – 0.9

• hs ≈ 10 x nhs

• large variability within hs

11 PAH compounds

0 5 10 15 20 25

04-01 05-01 06-01 07-01 08-01 08-31 10-01 11-01 12-01 01-01 01-31 03-02 04-01 05-02 06-02 07-02 08-02 09-01

P A H c o n ce n tr a ti o n ( n g m

Date in 2011-2012 (mm-dd)

2011 2012

Lipponen et al. 2013: Technical EXPAH Report on infiltration and

exposure model with software prototypes, 52 pp.

Seasonal and µE differences

Observations:

• All µEs similar

• winter >>> nhs

• mean i/o ratios 0.6 – 0.8

Lipponen et al. 2013: Technical EXPAH Report on infiltration and

exposure model with software prototypes, 52 pp.

Role of microenvironments

• Residential outdoor levels higher than indoors and outside vehicles (car, bus)

– traffic exposures do not seem high

seem high

– focus on infiltration

Overall observed average

Analysis of

Exposure determinants

• What are the main sources of exposures in Rome?

– Space heating likely dominating source (winter levels high)

=> Which microenvironments are critical for exposures?

• Traffic?

• Traffic?

• Outdoors?

• Indoors: Homes, schools, offices?

• indoor levels lower than outdoors:

– indoor sources limited

– infiltration important (may halve exposures)

– role of traffic not dramatic but cannot yet be excluded

Time-activity

• Large differences in time-activity between age groups:

– Elderly spend more time at home – and children at school (obviously)

– children spend slightly more time

n=483

(7-8 yr)

– children spend slightly more time outdoors but also slightly less walking and biking

n=707

(65-85 yr) Lipponen et al. 2013: Technical EXPAH Report on infiltration and

exposure model with software prototypes, 52 pp.

Based on data from Technical EXPAH Report on Time-activity

pattern of children and elderly in Rome. Action 3.1, 37 pp.

Aerosol Processes in PAH Infiltration and Population Exposure in Rome

P. Lipponen ¹ , O. Hänninen ¹ , R. Sorjamaa ¹ ,

M. Gherardi ² , M.P. Gatto ² , A. Gordiani ² , A. Cecinato ³ , P. Romagnoli ³ and C.

Gariazzo ²

1

National Institute for Health and Welfare (THL),

Italian Workers’ Compensation Authority (INAIL),

National Research

1

National Institute for Health and Welfare (THL),

Italian Workers’ Compensation Authority (INAIL),

National Research

Council of Italy, Institute of Atmospheric Pollution Research (CNR-IIA)

Infiltration:

mass-balance

(Eq. 1)

) (

)

( t a k

C k

a V C Q

k a

C _i Pa _a ⁱ

+

∆

− ∆ + +

= +

ig a

i F C C

C = _inf +

Wilson et al., 2000 Wilson & Brauer, 2006 Hänninen et al. 2004, 2013

k a

F Pa

= +

inf

P = penetration efficiency [0..1]

) (

)

( a k t a k

V k

a + + ∆ +

where

C

= indoor concentration (µgm

)

C

= ambient (outdoor) concentration (µgm

) C

= indoor generated concentration

P = penetration efficiency (dimensionless) a = air exchange rate (h

)

k = decay rate indoors (h

)

Q = source strength (µg h

) (symbol used by Dockery and Spengler was S) V = interior volume of the building (m

)

(Eq. 2) ai C a F C a

k a

C Pa = _inf

= + ^{(Eq. 3)}

) ( a k V

C _ig Q

= +

P = penetration efficiency [0..1]

a = air exchange rate [h ^-1 ]

k = decay rate [h ^-1 ]

Observations

Compound Outdoor measurements Indoor measurements

mean sd n mean sd n

Carcinogenic

1 BaP ng m

0.56 0.80 145 0.41 0.55 146

2 BaA ng m

0.31 0.40 145 0.17 0.22 146

3 BbF ng m

0.90 1.24 48 0.51 0.70 54

4 BjF ng m

0.09 0.02 24 0.08 0.03 24

5 BkF ng m

0.06 0.02 27 0.06 0.02 24

6 CH ng m

0.71 1.04 55 0.33 0.35 56

7 DBA ng m

0.10 0.14 130 0.07 0.08 132

8 IP ng m

0.62 0.84 145 0.46 0.57 146

Non-carcinogenic

9 BPE ng m

0.69 0.96 145 0.53 0.67 146

10 PE ng m

0.35 0.55 50 0.16 0.19 48

11 BeP ng m

0.68 1.00 55 0.55 0.64 56

Combined peaks

12 BbjF ng m

0.16 0.10 3

13 BjkF ng m

2.12 2.25 24 1.18 0.93 30

14 BbjkF ng m

0.91 1.29 94 0.59 0.65 92

Summed

cPAHs

ng m

3.35 2.08

Indoor and outdoor

measurements reported in EXPAH Techical

Report/Action 3.3

cPAHs ng m 3.35 2.08

ΣC×TEF

ng m

0.76 0.54

sPAH

ng m

7.55 13.46 11 2.34 3.59 12

ΣPAH

ng m

4.15 6.50 146 2.85 3.77 146

Ratios

BaP/ΣPAH % 13 % 5 % 145 14 % 5 % 146

BaP/PAHs % 5 % 1 % 2 9 % 2 % 5

BaP/BeP % 13 % 30 % 49 15 % 31 % 50

Particulate matter air pollution

PM25 µg m-3 22.6 13.1 143 22.5 15.6 140

EC µg m-3 2.2 1.3 28 1.8 1.5 28

OC µg m-3 6.7 6.0 28 6.6 3.0 28

Calculated sums

ΣBbjF ng m-3 0.31 0.82 146 0.20 0.49 146

ΣBbjkF ng m-3 0.32 0.82 146 0.21 0.49 146

ΣBjkF ng m-3 0.03 0.06 146 0.02 0.05 146

a sum of carcinogenic PAHs (1-8)

b sum of Concentration × TEF for carcinogenic PAHs (using Cal-EPA TEF factors)

c sum of BaP, IP, B(b,j,k)F

d sum of compounds (1-11) (including combined peaks when individual peaks were not observed)

Regression estimation of the mass-balance parameters

Traditional interpretation

ß ₀ = Cig (concentration by indoor sources) ß ₁ = Finf (infiltration factor)

y = 0.6121x + 0.0649 R² = 0.772 y = 0.5111x + 0.0173

R² = 0.7016 2.50

3.00 3.50 4.00 4.50

In d o o r le v e l (n g m

)

BaP

BaP

Definitions

Indoor sources:

Ci > Co

i/o = 1

ß1 was in many cases affected by the datapoints Cin > Cout

- e.g. here by 20%

⇒ rough filter applied:

points Cin > Cout were left out from ß1 estimation, but were included for ß0

0.00 0.50 1.00 1.50 2.00

0.00 1.00 2.00 3.00 4.00 5.00

In d o o r le v e l (n g m

Outdoor level (ng m

)

1:1

Ci≤Co (for ß1)

Ci>Co (in ß0)

Cig ≈ ß0(All),

Finf ≈ ß1(Filtered)

Confidence limits of the slope used for testing of statistical significance

2.0 2.5 3.0 3.5 4.0 4.5

In d o o r co n ce n tr at io n

BaP (ng m-3) Site=all, µEnv=all, Season=all (n=115+30) BaP 1:1

Ci≤Co (for ß1) Ci>Co (in ß0) y=0.56x + 0.05

95% Confidence limits:

0.0 0.5 1.0 1.5 2.0

0.0 1.0 2.0 3.0 4.0 5.0

In d o o r co n ce n tr at io n

Outdoor concentration

y=0.56x + 0.05 (R²=82%, n=115) Cig (ß0, all data) 95%CI (ß1=0.51) 95%CI (ß1=0.61)

Substantial interbuidling

i/o variability (~2x)

Potential Finf determinants were investigated:

• Microenvironment type (home, office, school, vehicle)

• Season (summer, autumn, winter, spring, heating/non)

• PAH compound (11 individual compounds, groups)

• PAH compound (11 individual compounds, groups)

Can statistically significant differences

be observed in the estimates?

I/O regression analysis

Cig or Finf different in different µEs?

Run set 1: by microenvironment Regression results for Ci<Co dataset Unfiltered dataset Ci<Co -dataset

Finf(Ci<Co dataset) Cig (Full dataset) p(ß0=0) p(Finf(µe)=Finf(All))

CompoundSite µEnv Season ß1 SE1 R2 n ß0 SE0 R2 n

1 PM2.5 all all all 0.603 0.051 66.2 % 74 2.137 1.732 57.3 % 139 0.109 . 0.500 .

2 PM2.5 all home all 0.647 0.066 78.3 % 29 4.382 2.260 53.1 % 53 0.026 ** 0.351 .

• Tool for infiltration factor regression analysis

– β1=Finf estimate (SE1=standard error for β1), β0=Cig estimate (SE0 = standard error for β0)

– Analysis by compounds and compound groups

– Unfiltered dataset = all data included, Ci<C0 = indoor sources removed from data

Indoor source? Finf different by µE?

2 PM2.5 all home all 0.647 0.066 78.3 % 29 4.382 2.260 53.1 % 53 0.026 ** 0.351 .

3 PM2.5 all school all 0.738 0.089 82.2 % 17 7.245 2.082 53.8 % 33 0.000 *** 0.167 .

4 PM2.5 all office all 0.244 0.124 21.7 % 16 10.951 2.516 14.9 % 24 0.000 *** 0.020 **

5 PM2.5 all traffic all 0.602 0.126 69.5 % 12 -2.864 6.359 65.3 % 29 0.326 . 0.498 .

6 PM2.5 bus all all 0.586 0.355 57.6 % 4 -11.011 5.821 89.0 % 16 0.029 ** 0.483 .

7 BaP all all all 0.561 0.024 82.3 % 115 0.066 0.027 77.1 % 145 0.008 *** 0.500 .

8 BaP all home all 0.550 0.036 85.3 % 43 0.056 0.032 79.3 % 53 0.042 ** 0.433 .

9 BaP all school all 0.629 0.057 82.8 % 27 -0.027 0.072 81.9 % 33 0.356 . 0.203 .

10 BaP all office all 0.479 0.060 75.4 % 23 0.095 0.080 71.6 % 26 0.118 . 0.165 .

11 BaP all traffic all 0.639 0.037 93.6 % 22 0.119 0.047 78.4 % 33 0.006 *** 0.102 .

12 BaP bus all all 0.934 0.027 99.2 % 12 -0.004 0.057 84.1 % 20 0.472 . 0.000 ***

13ΣPAH all all all 0.500 0.025 76.7 % 121 0.778 0.189 74.3 % 146 0.000 *** 0.500 .

14ΣPAH all home all 0.529 0.031 86.6 % 48 0.608 0.262 81.1 % 54 0.010 ** 0.305 .

15ΣPAH all school all 0.773 0.056 89.3 % 25 -0.189 0.378 89.7 % 33 0.309 . 0.000 ***

16ΣPAH all office all 0.305 0.040 72.9 % 24 1.216 0.423 70.8 % 26 0.002 *** 0.001 ***

17ΣPAH all traffic all 0.626 0.045 89.7 % 24 0.590 0.232 80.1 % 33 0.006 *** 0.038 **

18ΣPAH bus all all 0.913 0.036 98.2 % 14 0.022 0.266 87.4 % 20 0.467 . 0.000 ***

I/O regression analysis

Cig or Finf different by season?

Run set 2: home by seasons Regression results for Ci<Co dataset Unfiltered dataset Ci<Co -dataset

Finf(Ci<Co dataset) Cig (Full dataset) p(ß0=0) p(Finf(µe)=Finf(All))

CompoundSite µEnv Season ß1 SE1 R2 n ß0 SE0 R2 n

31 PM2.5 all home all 0.647 0.066 78.3 % 29 4.382 2.260 53.1 % 53 0.026 ** 0.500 .

32 PM2.5 all home heating 0.582 0.096 94.9 % 4 14.501 10.575 33.3 % 11 0.085 * 0.341 .

• Tool for infiltration factor regression analysis

– β1=Finf estimate (SE1=standard error for β1), β0=Cig estimate (SE0 = standard error for β0)

– Analysis by compounds and compound groups

– Unfiltered dataset = all data included, Ci<C0 = indoor sources removed from data

Indoor source? Finf different by µE?

32 PM2.5 all home heating 0.582 0.096 94.9 % 4 14.501 10.575 33.3 % 11 0.085 * 0.341 .

33 PM2.5 all home nonheating 0.324 0.068 49.8 % 25 12.272 2.138 5.8 % 42 0.000 *** 0.008 ***

33 PM2.5 all home winter 0.582 0.096 94.9 % 4 14.501 10.575 33.3 % 11 0.085 * 0.341 .

33 PM2.5 all home spring 0.617 0.097 67.9 % 21 8.899 2.325 16.9 % 34 0.000 *** 0.427 .

33 PM2.5 all home summer 0.216 0.113 64.8 % 4 18.227 5.707 0.2 % 8 0.001 *** 0.008 ***

33 PM2.5 all home fall

33 BaP all home all 0.550 0.036 85.3 % 43 0.056 0.032 79.3 % 53 0.042 ** 0.500 .

33 BaP all home heating 0.433 0.155 56.5 % 8 0.457 0.217 50.5 % 11 0.017 ** 0.269 .

33 BaP all home nonheating 0.638 0.050 83.3 % 35 0.016 0.007 79.0 % 42 0.011 ** 0.154 .

33 BaP all home winter 0.433 0.155 56.5 % 8 0.457 0.217 50.5 % 11 0.017 ** 0.269 .

33 BaP all home spring 0.640 0.057 81.7 % 30 0.015 0.009 78.2 % 34 0.049 ** 0.168 .

33 BaP all home summer 0.868 0.136 93.2 % 5 -0.005 0.018 67.9 % 8 0.389 . 0.032 **

33 BaP all home fall

33ΣPAH all home all 0.529 0.031 86.6 % 48 0.608 0.262 81.1 % 54 0.010 ** 0.500 .

33ΣPAH all home heating 0.348 0.099 58.1 % 11 4.899 1.410 52.9 % 12 0.000 *** 0.081 *

33ΣPAH all home nonheating 0.618 0.048 82.8 % 37 0.148 0.055 76.5 % 42 0.003 *** 0.128 .

33ΣPAH all home winter 0.348 0.099 58.1 % 11 4.899 1.410 52.9 % 12 0.000 *** 0.081 *

33ΣPAH all home spring 0.618 0.049 84.2 % 32 0.108 0.055 81.5 % 34 0.024 ** 0.131 .

33ΣPAH all home summer 0.727 0.068 97.5 % 5 0.242 0.148 73.4 % 8 0.051 * 0.022 **

33ΣPAH all home fall

I/O regression analysis

Differences by compound?

Run set 3: by compound Regression results for Ci<Co dataset Unfiltered dataset Ci<Co -dataset

Finf(Ci<Co dataset) Cig (Full dataset) p(ß0=0) p(Finf(µe)=Finf(All))

CompoundSite µEnv Season ß1 SE1 R2 n ß0 SE0 R2 n

52 BaA all all all 0.406 0.022 73.1 % 124 0.038 0.014 61.7 % 145 0.004 *** 0.001 ***

53 BaP all all all 0.561 0.024 82.3 % 115 0.066 0.027 77.1 % 145 0.008 *** 0.167 .

54 CH all all all 0.213 0.038 41.3 % 47 0.167 0.043 44.7 % 55 0.000 *** 0.000 ***

55 BPE all all all 0.608 0.021 88.3 % 111 0.082 0.028 83.8 % 145 0.002 *** 0.015 **

56 DBA all all all 0.245 0.042 29.0 % 87 0.043 0.007 28.5 % 130 0.000 *** 0.230 .

57 IP all all all 0.592 0.020 88.7 % 111 0.071 0.024 84.3 % 145 0.001 *** 0.062 *

58 PE all all all 0.249 0.036 58.9 % 36 0.061 0.020 63.4 % 47 0.001 *** 0.000 ***

59 BeP all all all 0.427 0.040 74.3 % 41 0.227 0.074 50.6 % 55 0.001 *** 0.000 ***

60 BbF all all all 0.531 0.043 79.4 % 42 0.042 0.059 80.2 % 48 0.236 . ***

Indoor source? Finf different by µE?

0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0

0.0 10.0 20.0 30.0 40.0 50.0

Indoor concentration

Outdoor concentration ΣPAH (ng m-3) Site=all, µEnv=all, Season=all (n=121+25)

ΣPAH 1:1 Ci≤Co (for ß1) Ci>Co (in ß0)

y=0.50x + 0.66 (R²=77%, n=121) Cig (ß0, all data) 95%CI (ß1=0.45) 95%CI (ß1=0.55)

61 BjF all all all 0.769 0.120 69.5 % 20 -0.005 0.019 46.8 % 24 0.407 . 0.005 ***

62 BkF all all all 0.728 0.107 74.2 % 18 0.021 0.009 42.3 % 24 0.008 *** 0.000 ***

63 BbjF all all all 0

64 BjkF all all all 0.242 0.067 44.7 % 18 0.758 0.207 36.8 % 24 0.000 *** 0.474 .

65 BbjkF all all all 0.655 0.033 84.9 % 73 0.055 0.041 80.0 % 91 0.087 * 0.050 *

66ΣBbjF all all all 0.526 0.043 78.5 % 43 0.035 0.020 79.8 % 146 0.036 ** 0.089 * 67ΣBbjkF all all all 0.523 0.045 77.5 % 42 0.038 0.020 79.7 % 146 0.028 ** 0.006 ***

68ΣBjkF all all all 0.914 0.086 84.4 % 23 0.000 0.001 91.9 % 146 0.373 . 0.001 ***

69 PAHs all all all 0.274 0.011 98.7 % 10 0.437 0.154 98.7 % 11 0.002 *** 0.000 ***

70ΣPAH all all all 0.500 0.025 76.7 % 121 0.778 0.189 74.3 % 146 0.000 *** 0.500 .

71 BaP/ΣPAHall all all 0.914 0.041 89.8 % 57 0.015 0.006 78.1 % 145 0.003 *** 0.000 ***

72 BaP/PAHs all all all 0

73 BaP/BeP all all all 0.542 0.100 61.8 % 20 0.069 0.041 27.2 % 49 0.045 ** 0.487 .

74 PM2.5 all all all 0.603 0.051 66.2 % 74 2.137 1.732 57.3 % 139 0.109 . 0.000 ***

75 EC all all all 0.717 0.080 79.2 % 23 -0.345 0.238 81.2 % 28 0.074 * 0.000 ***

76 OC all all all 0.537 0.040 95.8 % 10 3.545 0.370 82.1 % 28 0.000 *** 0.000 ***

Discussion on statistical analyses

• Existence of indoor sources

– high statistical significance seen but

– low magnitude => no practical relevance

• There is some (but limited) evidence on seasonal

• There is some (but limited) evidence on seasonal differences

– supported by earlier meta-analysis of PM2.5 across Europe (Hänninen et al., 2011)

• Suggestive evidence for differences between PAH compounds

Infiltration estimates PAHs

• Summer infiltration is higher than all season combined

• Weak evidence for difference for heating season and winter (lower infiltration as could be expected)

• Limited evidence in the data on

Microenvironm ent

Non-

heating Heating All

Homes 0.62 0.62 0.66

Schools 0.82 0.68 0.67

Offices 0.39 0.33 0.56

Season

• Limited evidence in the data on statistically significant seasonal differences in infiltration

• For PAHs the infiltration seems to be slightly higher during the non-heating season

Car and bus 0.4, 0.79 0.51, 0.88 0.59, 0.92

Walk and bike 1.00 1.00 1.00

Microenvironmental exposure model

∑ ^×

= f _i C _i E

Selection of relevant

microenvironments Corresponding time-activity data

E ₁ =f ₁ x C ₁ E ₂ =f ₂ x C ₂ E ₃ =f ₃ x C ₃ 1 Indoors (H, S, O)

2 ^Outdoors 3 In t ^raffic

Microenvironment Fraction of time Concentration Partial distribution distribution Exposure

0 10 20 30 40

Frequency

12 3510 20 3050 100 200

300500 Concentration [µg m ]-3

0 10 20 30 40

Frequency

12 3510 20 3050 100 200

300500 Concentration [µg m ]-3

0 500 1000

Frequency

12 3510 203050 100 200 300500 Concentration [µg m ]-3

x x x

Density

Fraction of time

0 0.2 0.4 0.6 0.8 1.0

0 2 4 6 8 10 12

Density

Fraction of time

0 0.2 0.4 0.6 0.8 1.0

0 2 4 6 8 10 12

Density

Fraction of time

0 0.2 0.4 0.6 0.8 1.0

0 2 4 6 8 10 12

Average personal exposure: Etot=E ₁ +E ₂ +E ₃

time-activity data

• Children and elderly PAH exposure was calculated by combining the time-activity data from questionnaires (483 children and 707 elderly), with average concentration levels in different microenvironments, indoors and outdoors

PAH Exposures

outdoors

• In both cases, winter season

strongly dominates the exposure.

In summer season people spend

more time outdoors than in winter

time, but PAH concentrations are

much lower in summertime.

Annual exposures

Contribution of microenvironments

Home indoors 67 %

Home outdoors

9 % Walk 12 %

Car 1 %

Bus 0 % Other

School, in 13 % Home, in

54 %

Home, out 4 %

Walk 4 %

Car 2 % Bus

0 % Other

17 %

• µE concentrations are rather similar

• => Exposure patterns driven by time-activity

Work indoors 1 % Work

outdoors 0 %

Other 10 %

Elderly annual ∑PAH exposure School, out

Children annual ∑PAH exposure 6 %

0 1 2 3 4 5

Elderly Children

Annual ∑PAH Exposure ng m-3

Other Bus Car Walk Home outdoors Home indoors Work outdoors Work indoors

Exposure modelling & analysis

Summary

• Biomass combustion proposed as the major source

• Winter exposures dominate (ca. 80% of annual)

• Indoor sources exist but are negligible

• Infiltration the single most important exposure modifier

• Minor differences between

• Minor differences between

– exposures of children and elderly – infiltration of PAH compounds

• Probabilistic, statistical and deterministic exposure models

developed and available for risk assessment and evaluation

of alternative policies

Exposure models available

Probabilistic model

Work indoors 1 % Work

outdoors 0 % Home indoors

74 %

Home outdoors

6 % Walk

8 % Car 3 % Bus Other 1 %

7 %

Elderly annual ∑PAH exposure

Microenvironmental model

Determinisitic spatial model

0 % 1 %

Elderly annual ∑PAH exposure

Greetings from Kuopio, FINLAND Greetings from Kuopio, FINLAND Greetings from Kuopio, FINLAND

Mille grazie – Saluti da Kuopio!

OTTO.HANNINEN THL.FI @ @