Cookstove Smoke Impact on Ambient Air Quality and Probable Consequences for Human Health in Rural Locations of Southern Nepal

Residential emission from traditional biomass cookstoves is a major source of indoor and outdoor air pollution in developing countries. However, exact quantification of the contribution of biomass cookstove emissions to outdoor air is still lacking. In order to address this gap, we designed a field study to estimate the emission factors of PM2.5 (particulate matter of less than 2.5 µ diameter) and BC (black carbon) indoors, from cookstove smoke using biomass fuel and with smoke escaping outdoors from the roof of the house. The field study was conducted in four randomly selected households in two rural locations of southern Nepal during April 2017. In addition, real-time measurement of ambient PM2.5 was performed for 20 days during the campaign in those two rural sites and one background location to quantify the contribution of cooking-related emissions to the ambient PM2.5. Emission factor estimates indicate that 66% of PM2.5 and 80% of BC emissions from biomass cookstoves directly escape into ambient air. During the cooking period, ambient PM2.5 concentrations in the rural sites were observed to be 37% higher than in the nearby background location. Based on the World Health Organization (WHO)’s AirQ+ model simulation, this 37% rise in ambient PM2.5 during cooking hours can lead to approximately 82 cases of annual premature deaths among the rural population of Chitwan district.


Grimm
The environmental dust monitor (Grimm EDM 180) is a standard stationary optical aerosolmonitoring instrument designed by Grimm Aerosol Technik GmbH & Co. Kg, Dorfstrasse-9, Germany. It can simultaneously measure the mass fraction of PM1, PM2.5, PM10 and TSP in 0.1 μg/m 3 resolution. Particle concentration are measured according to the principle of light scattering of a single particle at a constant air sample flow rate of 1.2 L/min. To protect the semi volatile loss of particle during sampling, an isothermal inlet system was made that uses the Nafion dryer. The Grimm is widely used by the scientific community for atmospheric research [1,[6][7][8]. In this study, a factory calibrated Grimm placed at Chitwan's air quality monitoring station (CAQMS) was used. Details about the instrument can be found at https://www.grimm-aerosol.com/productsen/environmental-dust-monitoring/approved-pm-monitor/edm180/.

Comparison between E-Sampler and Grimm
To correct the bias in reading of the E-Sampler, data obtained from it was compared with data obtained from the standard equipment, Grimm. A correction factor was derived by comparing the data and used to correct the E-Sampler data. Data from all five E-Samplers (two used in village ambient measurement, one used in background air measurement and two others used in indoor and outdoor measurement) were compared with data from the Grimm placed in CAQMS operating simultaneously. An example graph showing a comparison of data from the E-Sampler with data from the Grimm is presented in Figure S2. The variation in the observed concentration can be attributable the measurement process of the two instruments. Grimm is accompanied with Nefion dryer and thus avoiding any loss of volatile organics [9]. Mahapatra et al [1] also stated about the underestimation PM2.5 by esampler due to loss of volatiles during the measurement compare to Grimm. Hence the difference in the concentration is expected between Grimm and Esampler. The correction factors of all other E-Samplers are given in Table S1.

MicroAeth
Black carbon mass concentration was measured using the microAeth model [10]. The instrument is portable and siphons air at a constant flow rate of 50 mL/min. It measures the BC concentration based on the light absorption principle at the 880 nm wavelength of light. The inlet of the microAeth was fitted with a PM2.5 cyclone in order to prevent particles with a larger diameter (greater than 2.5 micrometers) from getting inside the equipment. The microAeths used in the field were factory calibrated to ensure high-quality measurements.

Aethalometer
Aethalometer AE33 (Magee Scientific, USA) is a standard aerosol measuring equipment, used mainly for black carbon and brown carbon measurement. AE33 collects the air pollution sample and analyses it in seven optical wavelengths ranging from the near-infrared (950 nm) to the near-ultraviolet (370 nm) and provides real-time data. It measures the attenuation (ATN) of a light spectrum passing through a filter on which aerosols are continuously collected. Data obtained from the 880 nm channel is generally taken as black carbon reading.

Comparison of MicroAeth and Aethalometer
MicroAeths used in indoor and outdoor emission measurement were operated simultaneously with the aethalometer and the data obtained from all three equipment were compared. The scatter diagrams showing a comparison of data from the microAeth used in both indoor and outdoor smoke measurement are presented in Figure S3 and the correction factors are given in Table S2.

Indoor Air Quality (IAQ) Probe
The IAQ Probe was used to measure the CO2 and CO concentration. It is a widely used, reliable equipment for measuring indoor air quality [11,12]. These probes were also compared against highquality calibration gas standards (Specialty Gases Ltd. and Alchemic Gases and Chemical Pvt. Ltd., Mumbai, India) and subsequently correction factors were derived. The correction factor of the CO2 measurement was calculated by comparing the data from sensor against standard CO2 calibration gas mixture concentration at 350 ppm and 1200 ppm (Specialty Gases Ltd.). Similarly the correction factor for CO sensor was also calculated by comparing the data from sensor against 5 ppm (Specialty Gases Ltd.) and 81 ppm (Alchemic Gases and Chemical Pvt. Ltd., Mumbai, India) CO calibration gas mixture.

S4: Emission factor calculation equation,
The total carbon (TC) in the cookstove smoke was calculated by the summation of carbon content in CO2 (cCO2), CO (cCO), BC. The emission factor (EF) of CO (EFco) was then computed as in Eqn (1): Where Fc is the carbon weight fraction taken as 0.5 and mCO is the carbon fraction in CO. The computed EF was expressed in grams of pollutant emitted per unit kg of fuel wood combusted.
Similarly, the EF for CO2 was computed as in Eqn 2: The EF for particulate pollutants such as PM2.5 and BC were measured relative to EFCO following Eqn 3: Where, Mx is the mass concentration in μg/m 3 of particulate pollutants such as PM2.5 and BC.