Utility of low cost CO sensing: measuring exposure in Ghana to source apportionment of other
pollutants Mike HanniganUniversity of Colorado
REACCTINGResearch of Emissions, Air Quality, Climate, and Cooking Technologies in Northern Ghana
Intervention Study
rural households (200)
Sources and Exposure Characterization
urban and rural regions(50 households)
2
Electrochemical CO sensor
How did calibration go?Calibration Process
• Pre-deployment lab calibration• In-field concentration check via colocation• Post-deployment lab calibration
Biggest issue (on average) seen with calibration was the discrepancy between factory and pre-deployment
LascarSignal = p1 + p2(ReferenceConcentration) + ε
The average slope (p2, sensitivity) was 1.06 ± 0.06 and intercept (p1) was 0.08 ± 0.13 ppm
Pay attention to [CO] scale
Piedrahita et al., Exposures to Carbon Monoxide in a Cookstove Intervention in Northern Ghana, Atmosphere, in press.
How well does it work in practice?
0 5 10 15
CO ppm
0
2
4
6
8
10
12
CO
ppm
dup
licat
e
Calibrated mean exposure duplicates
R = 0.78013, ARPD = 237.3576,RMSE = 1.1622
0
50
100
150
200
250
300
350
400
450
500Days since first deployment
0 2 4 6 8 10 12 14
CO ppm
0
2
4
6
8
10
12
CO
ppm
dup
licat
e
Raw mean exposure duplicates
R = 0.91623, ARPD = 431.6972,RMSE = 0.69136
0
50
100
150
200
250
300
350
400
450
500Days since first deployment
We collected 141 days of duplicate CO exposures.
Performance over time
We used 34 units.Goals was 2 years of use.Issues …
• Batteries• Sensors
Example CO personal exposure
6
Personal CO Doses
G/P 3-stone Female Male Over 5 yr Under 5 yr
7
P/PG/G
)rh 8 4re
pnm
i-
mpp (esoD
Personal CO Doses
G/PP/PG/G 3-stone Female Male Over 5 yr Under 5 yr
Dose
(ppm
-min
per
48
hr)
8Why aren’t we seeing differences in CO exposure?
Personal CO Doses
G/PP/PG/G 3-stone Female Male Over 5 yr Under 5 yr
Dose
(ppm
-min
per
48
hr)
9Why aren’t we seeing differences in CO exposure?
Is there no difference in emissions between stoves in real world?Are intervention stoves displacing traditional stoves?Are other CO sources driving exposure?
Is there a different way to check this exposure – source link?
Use real-time exposure data with proximity & location data.
Low-cost real-time sensing can provide unique opportunities for analysis.
Piedrahita, et al., Attributing Air Pollutant Exposure to Emission Sources with Proximity Sensing, Atmosphere, in press.
Sensor time resolution helps find links with sources and activities …
Can the CO sensor help with PM measurements?
E. Coffey et al., “Kitchen Area Air Quality Measurements in Northern Ghana: Evaluating the Performance of a Low-Cost Particulate Sensor within a Household Energy Study”, Atmosphere, in press.
MCE to improves PM sensor quantificationSo does temperature, RH, season, urban/rural classification
How useful is this type of CO sensor at lower US ambient concentrations?
Here is a co-location of 5 Pods (equipped with electrochemical CO) in LA
You can use them to explore source links…
Collier-Oxandale, A, et al. “Using Gas-Phase Air Quality Sensors to Disentangle Potential Sources in a Los Angeles Neighborhood”Submitted to Atmospheric Environment
You can use them to explore source links…
This work was done by many people …
My group:Evan Coffey, Ashley Collier-Oxandale, Joanna Casey,Jake Thorson, Ricardo Piedrahita, Drew Meyers, Nick Masson
Collaborators in SoCal:Jill Johnston (USC)Sandy Navarro (Esperanza Community Housing)Many people at SCAQMD, CARB
REACCTING teamChristine Wiedinmery (CU)Katherine Dickinson (Colorado School of Public Health)Maxwell Dalaba, Ernest Kayomanse, and Abraham Oduro (NavrongoHealth Research Centre)
Funding:Environmental Protection Agency STAR Program (#83542401)
National Science Foundation MetaSense (#1446912)National Science Foundation AirWaterGas SRN (#1240584)
National Science Foundation CyberSEES (#1442971)National Institutes of Environmental Health Sciences (#R21ES027695)
Top Related