Rural and urban exposure to indoor air pollution

1
Rural and urban exposure to indoor air pollution

• Sumeet Saksena
• East West Center
• Honolulu

2
Alternative title Satisfying the curiosity of
the ambient air pollution experts about IAP

• Principle used do not preach to the converted
• What do we know and what do we not know about IAP
  related human exposures?
• Why we do what we do?

3
Basics of human exposure assessment

• Whose exposure?
• Infants, children, adults (women?), elders, etc.
• Influenced by health outcome focus
• Implications on protocols (breathing zone,
  activity patterns, etc)
• Where is the exposure happening?
• Indoors (kitchen, living room), Outdoors (yard,
  near house, far from house)
• Duration of exposure?
• High concentration short duration exposure
• Low concentration long duration exposure

4
Exposure assessment direct approach

• Personal monitoring for 24 hours or lesser

5
(No Transcript)
6
Exposure assessment indirect approach
micro-environmental modeling

• Identify major microenvironments
• Indoors during cooking
• Indoors non-cooking (day vs. night)
• Outdoors (near ambient)
• Daily exposure is time weighted average of area
  levels (breathing height) in these
  microenvironments

7
(No Transcript)
8
(No Transcript)
9
Typical levels of key pollutants related to
cook-stoves (wood fuel)
10
Typical levels of key pollutants related to
cook-stoves (kerosene/gas)
11
Typical levels of key pollutants related to
cook-stoves key messages

• Anomalies partly due to lack of uniformity in
  measurement protocols
• Background PM levels are high

12
Particulate matter size distribution

• Recent study in Costa Rica indicated two peaks at
  0.7 and 2.5 microns
• In lab studies, unimodal aerosol size
  distributions observed with mass median
  aerodynamic diameters of 0.5-0.8 microns

13
Trends in measurement methods

• PM gravimetric
• Low-flow pumps for area/stationary or personal
  sampling
• Medium-flow pumps for area/micro-environmental
  sampling
• Cyclones for 3.5, 4, and 5 microns
• Impactors for 2.5 and 10 microns
• Advantage standard methods available (e.g. NIOSH
  0600), further chemical analysis
• Disadvantage cost ( 1500 per kit), high QA/QC
  skills, electricity in the field, etc.

14
Trends in measurement methods (cont.)

• PM real time
• Based on optical scattering/ ionization. Very few
  studies so far
• Advantage real time data, some makes are
  inexpensive (UCB monitor), multi-stage size
  cut-offs
• Disadvantages commercial access difficult, some
  models cannot be used in personal mode, particle
  size-cut off convention different from
  traditional conventions.

15
Trends in measurement methods (cont.)

• CO
• Potentiometric dosimeters standard, durable,
  expensive
• Diffusion tubes cheap, 25 error

16
Spatial variations

• Why study?
• Whose exposure?
• Defining the breathing zone
• Within the kitchen variations
• Horizontal distance from stove, but being far
  is not necessarily safer depends on fuel, stove
  and ventilation conditions
• Vertical variations smoke hangs at about 4 feet
• Inter-room variations
• Cook moves around
• Others in the family

17
Temporal variations

• Across meals
• Day-to-day variations mixed results so far
• Seasonal variations few studies, changes in type
  of fuel, cooking activity, weather and
  ventilation
• Few studies conducted with long sampling duration
  (e.g. one week), but short term measurements not
  made simultaneously, so cannot conclude

18
Impacts of stove interventions

• Type 1 studies cross-sectional designs
• Early studies in India and Nepal led to
  inconclusive results due to design weaknesses
  (confounding)
• Modest benefits for TSP, better for CO
• Hoods more effective
• Problems due to neighbours smoke and background
  levels
• Type 2 studies before-and-after comparisons
• PM reductions of 40-60
• Area levels reduction gt personal levels reduction
• Study in Kenya found hoods far more effective,
  and windows ineffective

19
Correlations between and among pollutants

• Why study this?
• Identify simple and inexpensive proxy indicators
  for PM
• Simplify personal monitoring
• Correlation between CO and PM (co-located
  sampling)
• Degree of correlation higher over longer periods
  of time as compared to shorter periods
• As PM size decreases correlation with CO
  increases
• Mixed results across studies. Degree of
  correlation depends on stove, fuel, ventilation
  factors, etc.

20
Correlations between and among pollutants (cont.)

• Correlation between area and personal sampling
• Mixed results across studies
• Situation specific

21
Other major explanatory factors

• Recent studies have provided evidence of the
  important role of
• Type of house
• Location of kitchen
• Kitchen architecture
• Ventilation
• There is an urgent need to have standard
  definitions for the above parameters (e.g. what
  is open cooking?)

22
Role of time activity patterns

• Obvious more time spent cooking greater the
  exposure
• Not so obvious Interventions and natural
  transitions not only change emissions but may
  impact on activity patters and behaviour. The NET
  impact on exposure can be positive, negative or
  zero.

23
Exposure across fuel groups Delhi slum case study

• Mean daily exposure not significantly different
  between wood and kerosene users
• More meals, more items, longer meals longer
  cooking times in kerosene houses
• Kerosene users cook indoors, wood users outdoors
• Infants of kerosene users near stove longer

24
Exposures in urban/peri-urban areas

• Concentration levels during cooking same as in
  rural areas same fuels, stoves, small kitchens
• Exposure to PM due to cooking as a fraction of
  daily exposure
• gt 75 in rural areas
• 10-20 in urban areas

25
Exposures in urban/peri-urban areas (cont.)

• Dense housing implies
• Smoke from one house infiltrates another house
• High near-ambient levels (gt 500 ug/m3 for PM5)
• Need community-wide interventions
• Cluster of houses act as an area source of
  ambient air pollution at micro-urban scales
• Indoor-outdoor relationships are complicated and
  not well studied

26
IAP exposures and epidemiology

• Only one study (Kenya) quantified the link
  between exposure and incidence of ALRI
• Concave relationship
• Rate of increase declining above 11-2 mg/m3
• Highlighted the role of short term peaks
• Elevated levels occur during fire ignition
  (especially for coal and charcoal) and fire
  tending
• There is a need to have standard sampling
  durations (15 minutes, meal time, 24 hour?)

27
Key research questions/issues for the future

• PM size distribution under field conditions
• Correlation between area and personal sampling
• Quantifying the impact of housing and ventilation
  variables
• Indoor-outdoor relationships in urban/peri-urban
  areas
• Measurement of acute exposures

28
Key issues for aid agencies for the future

• Development of simple and inexpensive methods and
  protocols
• Identification of other types of interventions in
  addition to improved stoves
• Harmonization of methods
• Training of NGO trainers
• Creating repositories of instruments, Technical
  Backup Units (linked to NGOs) with advanced
  infrastructure