Particulate Matter and Carbon Monoxide Emission Factors from Incense Burning

.......................................................................................................................................... x


Introduction
Indoor air quality (IAQ) has become a major concern in the present-day society and has triggered the interests of many researchers around the globe. The World Health Organization states that there are millions of deaths every year attributed to indoor air pollution in the world [1] . In the modern world, most people spend 90% of their time living indoors [2] . Indoor sources like cooking, charcoal burning, incense, smoking, household cleaners, and candle burning contribute to significant amounts of indoor air pollution. People upon exposure to this kind of pollution are prone to various health risks.
Indoor air pollution has been a major problem in many developing countries. The use of coal, and wood as a source of energy for cooking, lead to increased smoke, fine particulate dust, and other pollutant emissions due to the incomplete combustion. About 3 billion people, mostly from developing nations, use biomass as a source of energy to cook food and heat their homes [3] . The use of biomass generates pollutants that can accumulate to hazard levels in a short time interval, which can lead to adverse health effects. A WHO report states that there were 4.3 million deaths attributed to the household air pollution from cooking with solid fuels in 2012 [3] . Women and children are most exposed to the indoor air pollution as they spend most of their time at homes.
Children breathe more air per kilogram of body weight than adults do [2] . Thus, children are more prone to indoor air pollution.
Incense burning is a common tradition followed in many Asian countries; this activity is mainly used for religious and aesthetic purposes. Incense sticks are also used at homes to mask bad odors, repel mosquitoes, flies, and insects. Indoor air pollution resulting from burning incense is a major public health concern. Excessive usage of these incense sticks releases toxic pollutants which are leading to adverse health effects and sometimes death too. The amount and kind of pollutants released from the incense depend on the composition of the material used in making the incense sticks. Generally, incense burning generates the respirable particulate dust, carbon dioxide, carbon monoxide, volatile organic compounds (VOCs) and others.
Incense sticks are widely used in temples in many Asian countries as shown in Figure 1. The smoke produced pollutes air both in and around the temples, thus people are more exposed to the harmful pollutants for several hours a day. A report by the Environmental Protection Agency in Taiwan in 2003 stated that a total of 28.7 metric tons of incense was burned in 92 temples in Kao-Hsiong city [4] which emits significant quantities of air pollutants. The increasing levels of smoke generated is jeopardizing the health and lives of people. The purpose of this research is to develop a lab scale model to estimate the PM2.5 (number and mass), PM10 (number and mass), and CO emission factors (EFs) from incense burning. EFs indicate how polluting a specific activity is. It is an amount of air pollutant emitted per unit amount of work done, or product consumed, or product produced, or something similar. Thus, units of EFs in this case can be "g/g", but for convenience, they can be expressed as "g/Kg", or "mg/g" and so on. The EFs generated in this research will be helpful to many researchers, scientists to develop emission control strategies, health risk assessments.
Source: Google Images

Ojectives
The main objective of this research is to estimate the emission factors for the following pollutants from incense stick burning • Carbon monoxide (CO) • PM2.5 (mass) • PM2.5 (number) • PM10 (mass) • PM10 (number) The above objectives were pursued using the following tasks: •

Emission Factor
The amount/quantity of a pollutant released to the atmosphere from a specific process or activity is known as Emission Factor (EF). It is expressed as the amount of pollutant emitted per unit amount of work done [5] . (Or) The emission factor is an indicator of how polluting a specific activity is. It is an amount of air  [5] . EFs are used in developing greenhouse gas (GHG) emission inventories for ports, steel industry, electric generation utilities and so on, and they will help in creating awareness among people and helps in developing emission control strategies to minimize emissions. EFs are used to estimate emissions from different sources, as well as in air permitting applications, environmental impact assessments, and public health risk assessments.  [5] . EF's can also be found in many scientific articles, publications of trade organizations such as National Shipbuilding Research Program (NSRP), and others. Industries can use these published emissions factors for emission inventories, air permit applications, environmental impact assessment, and health risk assessments.

Previous research conducted by UNO researchers on Emission Factors
Researchers in academic institutions typically simulate industrial/manufacturing processes in emissions testing facilities where they have a controlled environment allowing them to measure (a) activity (amount of work done; the amount of raw materials used; the amount of product produced), and (b) quantities of pollutants generated by various process conditions.
The UNO researchers [6,7,8]  Dry abrasive blasting is a surface preparation method that can be used on steel or other metal surfaces, for creating a rough profile, and for removal of contaminants like rust from the surface.
This process is used in many industries such as aerospace, automobile, metal finishing, shipbuilding and ship repair for preparing and maintenance of steel or other metal surfaces.
Abrasive materials are forced at high velocities on the surface with the help of compressed air for the preparation of the surface. This process results in the removal of contaminants and creating the rough profile on the surface. Coal slag, copper slag, silica sand, steel grit, steel shot, aluminum oxide and so on. are the most commonly used abrasives.
Dry abrasive blasting process emits particulate matter, metals, and spent blast media that harm the human health and environment. The commonly used abrasive material used for abrasive blasting is silica sand because of its low cost and abundant occurrence in nature. The emissions from silica sand are of greatest concern because it leads to adverse health effects on workers.
The emissions test facility was developed to conduct research on determining the particulate matter emission factors from dry abrasive blasting. This test facility was equipped with test chamber, blasting equipment, test plates, exhaust duct, stack sampling system, and particulate collection system as shown in Figure 2. Dry abrasive blasting was performed on the base plates in an enclosed test chamber using three different expendable abrasives: coal slag, copper slag, and specialty sand. The experiments were performed by varying blast pressure and abrasive feed rate to determine EFs under varying conditions. The pressure was varied using an air compressor, and the feed rate was varied using the blast pot. The dust produced from the process was drawn out from the chamber through exhaust duct using a fan. The particulate dust produced was collected on a filter paper in the filter box, and the dust produced was collected by the 2-stage particulate collection system as shown in Figure 2.
The EFs for various process conditions were calculated using equations developed. Two types of moreover, the EFs developed will be helpful in estimating the TPM emissions from dry abrasive blasting process and to develop emission control strategies to minimize emissions.
UNO research on emission factors (kg/m 2 , kg/kg) for the three abrasive materials indicate the following hierarchy based on the TPM EFs: Copper Slag < Coal Slag < Specialty Sand.
3.1.1.2 Heavy metal (Cr(VI) and total metals) emission factors from welding processes The UNO researchers [7,8] had conducted research on estimating heavy metal EFs from welding processes. This research was attempted to capture and analyze 100% of the fume generated from the electrodes; to serve the purpose, a weld fume chamber was designed as per American Welding Society (AWS) as shown in Figure 3. The purpose of the research was to estimate how many grams of Cr(VI) and total metals are emitted per 1g of weld rods (electrodes) consumed. The welding process was performed in the weld fume chamber and the fume generated was captured on the filter media (eight inches in diameter) installed on the top opening of the fume chamber as shown in Figure 3. The OSHA ID-215 and the NIOSH 7300 methods were selected for the analysis of Cr (VI) and total metals, respectively in this research.
The experiments were designed for Shielded Metal Arc Welding (SMAW) and Flux Core Arc Welding (FCAW) welding processes using different electrodes.

Incense Sticks
Incense sticks are made of an aromatic material that produces fragrant smoke when burned.
Incense is available in various forms like sticks, joss sticks, cones, coils, powders, rope, rocks/charcoal, and smudge bundles as shown in Figure 4 [9] . Incense sticks are commonly used for aesthetic and religious purposes, to produce a pleasant smell, or mask bad odors in indoor environments. Typically, the composition of incense stick consists of 21% (by weight) of herbal and wood powder, 35% of fragrance material, 11% of adhesive powder, and 33% of bamboo stick [4] .
In many developing countries, incense is burned in homes and public places such as stores, shopping malls, and temples. They have been widely used in many Asian countries like India, China, Japan for various purposes. The incense sticks market in India is worth 225 million USD per year, and it is growing 10% annually. In China, 76.9% of incense is burnt at homes every day, and around 90% of the population has been using incense sticks for over 20 years [10] . volatile organic compounds (VOCs) and others. Pollutants released from incense burning accumulate quickly to hazardous levels in indoor environments, particularly in a poorly ventilated area [11] . The rise of pollutant levels in the indoor environment have adverse effects on human health.
Incense stick comes under combustion process. Combustion is the reaction between a fuel and oxidant, the fuel is carbonaceous such as gasoline, wood, or coal, and the oxidant is the oxygen present in the air. Typically, incense stick when burnt release particulate matter, carbon monoxide (CO), carbon dioxide and other pollutants such as sulfur compounds (if sulfur is present), oxides of nitrogen, organic compounds/hydro carbons (as most combustible material is made of organic material) [19] .
Source: Jetter JJ, et al., 2002 Chemical reactions are as follows:

Indoor Air Pollution
Indoor air quality has become an important topic in recent years. Nowadays, people spend about 80% of their time in indoor environments such as homes, workplaces, shopping malls, stores and so on [12] . Indoor sources like charcoal, incense, and candle burning, cooking, household cleaners, smoking tobacco products are contributing a significant amount of pollution in indoor areas. Hence people are being exposed to particulate dust, various gaseous pollutants, and VOCs and are prone to various health risks.
In 2012, there were 3.3 million deaths attributed to indoor air pollution in the low-and middleincome countries in the South-East Asia and Western Pacific Regions. The indoor air pollution related deaths include 34% due to stroke, 26% due to ischaemic heart disease, 22% due to chronic obstructive pulmonary disease (COPD), 12% due to pneumonia, and 6% due to lung cancer [1] . Indoor pollutants have a significant impact on human health, as people spend most of their time in indoor environments. The indoor air pollution leads to various health risks and death over prolonged exposure. In the most developing countries, people use charcoal, and wood as a source of energy for cooking, heating their homes. This process results in the production of particulate dust, various gaseous pollutant emissions because of incomplete combustion. Women and children are most exposed to indoor air pollution as they spend most of their time at homes.
The U.S. EPA states that many pollutants in an indoor environment can be 2-5 times, and occasionally more than 100 times higher than ambient levels [2] . Indoor air pollution has become a major public health concern; it has attracted many researchers over the past few years. There have been many studies on emissions and characterization of emissions from various sources in indoor environments.
There are several factors that result in the increased pollutant levels in indoor environments. Air particulates are generally categorized by how deep they can penetrate the human respiratory system. Particles that are less than 2.5 microns (PM2.5) in size are known as fine particles, and these pose the largest health risks because they can enter deeper parts into the human respiratory system and lead to adverse health effects. Particles that are greater than 2.5 microns and less than 10 microns (PM10) in size are known as coarse particles; these are too large to enter the human respiratory system, but prolonged exposures of PM10 can lead to respiratory illness.
Incense burning leads to the production of particulate matter, out of which PM2.5 is the most released pollutant, which is more hazardous and causes adverse health effects. There have not been any scientific articles on the health effects directly caused by the particles from incense burning [4] . The general health effects caused by the particle matter (PM) are summarized below.
Globally, 3% of cardiopulmonary and 5% of lung cancer deaths are attributable to PM. The health effects of inhalable PM due to short-term exposure include respiratory and cardiovascular morbidity, such as aggravation of asthma, respiratory symptoms and an increase in hospital admissions; long-term exposure risks include mortality from cardiovascular and respiratory diseases and lung cancer [13] .
Short-term exposure to PM10 has significant effects on respiratory health, but for mortality, PM2.5 is a larger risk factor than PM10 for prolonged exposures. Studies have shown that with a 10 μg/m 3 increase in indoor PM10 concentration, the cardiovascular mortality increased by 0.36% and respiratory mortality by 0.42%. Likewise, with a 10 μg/m 3 increase in indoor PM2.5 concentration, the cardiovascular mortality increased by 0.63% and respiratory mortality by 0.75% [10] .

Carbon Monoxide health effects
Carbon monoxide (CO) is a colorless, odorless, tasteless, and poisonous gas generally produced from the incomplete combustion of fossil fuels, burning of wood, incense, and tobacco. The inhalation of CO leads to the displacement of oxygen in the blood and deprives the heart, brain, and other vital organs of oxygen.
CO readily combines with hemoglobin much more than oxygen with a factor of 200-300 to form carboxyhemoglobin, this reduces the blood's capacity to transport oxygen. CO, when inhaled in low concentrations, leads to headaches, dizziness, weakness, and nausea, while high concentrations can be detrimental [4] .
Despite the evidence that incense burning leads to adverse health effects, still many people in the developing countries use them in their daily lives.

Previous studies
There were few previous efforts to evaluate the emissions rates and emission characteristics from incense burning in a controlled environment/laboratory setting. Some of them are summarized below: A study on the emission levels of particles (PM10) and benzene from incense burning, sparklers, and cigarettes in an indoor environment was conducted by Tirler in 2015. In his research, experiments were performed in a 75m 3 volume room. An aerosol spectrometer (Grimm model 1.108) was used for measuring the PM10 mass and number concentrations. The average background PM10 mass concentration in the room was 6 µg/m 3 . The experiments were performed using different kinds of incense sticks, and the PM10 mass concentrations range between 200 and 300 µg/m 3 . The instrument used in the experiment was expensive. Emission rate, flow rate, and emission factor computations were lacking in the research [11] . There was no repeatability in the experiments [14] . The emission rates of CO from ten different incense sticks ranged between 1.9-794.7 mg/hr, PM2.5 emission rates ranged between 9.8-2160.3 mg/hr, and PM10 ranged between 10.8-2536.6 mg/hr. The EFs of CO from ten different incense sticks ranged between 1.0-227.7 mg/g of incense, PM2.5 EFs ranged between 3.3-87.8 mg/g of incense, and PM10 EFs ranged between 4.3-123.6 mg/g of incense. The instruments used in the experiments were expensive, there was no exhaust system installed in the chamber, and the relation between the EF, emission rate and air flow was not reported [16] . The emission rates and emission factors for CO were not reported. The experiments were not conducted in a closed chamber [17] .
The literature review and discussion presented in the above sections lead to the following conclusions: • Particulates particularly PM10 and PM2.5 is a health concern whether it is outdoors or indoors, particularly indoor environment leads to increased exposures due to lack of dispersion.
• Similarly, carbon monoxide also leads to health concerns.
• Incense sticks and other forms of incenses are used around the world for a variety of purposes and particularly in worshipping at churches, temples, and other religious places where significant number of people gather in indoor environments, which becomes hazardous quickly due to lack of dispersion of released pollutants from incense burning.
• Emission factors are useful for understanding, planning, and managing health risks associated with air pollution sources.
• There were limited attempts in the past to understand the incense burning and the associated emission factors for PM10, PM2.5, and carbon monoxide. Also, the prior studies have several limitations.
• Thus, this proposed research to understand air pollution from incense burning and determination of emission factors for PM10, PM2.5, and carbon monoxide is considered beneficial to address the air pollution problems associated with incense burning.

Theoretical Analysis
A simple mass balance equation is developed to explore the behavior of indoor air pollutants as a function of infiltration of outdoor air, indoor sources and sinks, and leakage to the outdoor air.
Assuming the contents in the indoor environment are uniformly/well mixed, the following materials balance is presented [18] : The general solution for Eq (1) is Where, Ct = concentration at time t, g/m 3 Co = Initial Indoor concentration, g/m 3

Steady-state condition
In the steady-state condition, the concentration of the pollutant does not change with time ( = 0). By applying this principle, equation (1) In a special case, when the pollutant in the box is conservative (no decay of pollutant), the ambient concentration outside the box is zero (Ca=0), and the initial indoor concentration is zero (Co=0), then Eq (1) reduces to The above discussed indoor air quality model was used to attain the objective of this research.

Assumptions
The following assumptions were made for estimating the particulate matter and carbon monoxide emission factors from incense burning.
• The contents in the box are uniformly/well mixed • The ambient concentration is zero (Ca=0) • The initial indoor concentration is zero (Co=0) Particulate matter (PM) is removed through precipitation and/or absorption and/or adsorption with physical structure in the real world. As we do not have precipitation and the only obstruction is the walls of the chamber and the exhaust duct, these were considered negligible.
Thus, the pollutant decay rate for PM assumed to be negligible.
A study indicates that the CO decay rate constant is 0.0 s -1 [20] , so the CO pollutant decay rate is considered negligible.

Methodology
This section describes the methodology to compute particulate matter and carbon monoxide emission factors from incense burning. The indoor air quality model that was discussed earlier was used in computing the emission factors.
When the incense stick starts burning, the pollutant concentration inside the box will increase to some extent. By assuming the contents in the box are uniformly/well mixed, the concentration remains the same (steady-state condition) for some time, then it decreases and eventually becomes zero. This is the basic principle applied in this research. The following are the known and unknown parameters in the research: The known variables in the experiment include: • Volume of the box (V), m 3 • Concentration of pollutant inside the box at any given time (Ct), g/m 3 • Given time (t), sec The unknown variables in the experiment include: • Emission rate (E) -"µg/sec" or "mg/sec" E = Mass of pollutant released/time • Flow rate (Q) -"m 3 /sec" or "l/sec" The experiments were conducted in a test chamber designed for this research, the pictures of the chamber were presented in the materials section. The test chamber has a rectangular duct where the incense smoke flows in, and the sensors were installed at the 8*equivalent diameter distance to avoid any turbulence effect.
The equivalent diameter for rectangular duct was calculated using the formula: the CO and particulate matter sensors were installed at this distance to get the concentration readings during the experiment.

Estimating the flow rate in the exhaust duct of the test chamber:
EPA Method 1 provides guidelines for selection of sampling locations and traverse points for measuring velocity in ducts [21] . Thus, this method was employed in estimating the average velocity in the exhaust duct. As the cross-sectional area of the exhaust duct was known, the exhaust flow rate was computed by multiplying the average velocity with the cross-sectional area.
The velocity measurements were taken at the sampling location. The average velocity was computed in the exhaust duct as per the EPA Method 1 guidelines. A 3*3 matrix layout was taken as shown in Figure 7 and the velocity was measured at 9 traverse points (the centroid of each rectangle on the cross-sectional area of the exhaust duct) as shown in Figure 7. The readings were averaged to get the average velocity reading inside the exhaust duct. Where Q = flow rate, m 3 /sec A = area of cross-section, m 2 V = average velocity in the rectangular duct, m/sec Then, the emission rate was calculated using Eq (4) as discussed earlier with known flow rate and the average concentration reading from the sensors.
Note: C in the Eq (4) is Cavg in the experiment, steady-state average concentration. It is discussed briefly in the later sections.

Emission factor (EF) estimation:
Emission factor (EF) in this experiment is calculated using the formula given below Emission rate of pollutant (E) is computed using the method discussed earlier Incense stick burnt rate is calculated using the formula below =

Example case
An example case is described below showing how the emission factors and other variables involved in the experiments were computed.
The experimental data was taken from the sensors and organized in a spreadsheet that has two columns with date/time and the sensor readings. A graph was plotted with time on X-axis and sensor readings on Y-axis. Looking at the graph, some of the fields in the Excel spreadsheet were marked as special fields to indicate the start time, first peak, end peak, final time, and stop of the experiments; they are described below: Start time: Represents the time at which the experiment started Note: The total time used in Eq (9) is the difference between start time and final time.
Detailed description for computing (PM2.5/PM10, number and mass) and CO EFs are described below:

Carbon monoxide EFs:
The following are the steps involved in computing the CO EFs and other variables for each experiment: • First, the CO readings were downloaded from the CO sensor and saved in an Excel  • From the graph, first peak, end peak, start time, and final time were noted • The average concentration was computed between the first and end peak as shown in Figure 8.
• With the calculated flow rate and the average concentration, the emission rate (E) was calculated using Eq (4).
• Using the emission rate (E) calculated in the previous step, the emission factor was calculated using Eq (8) and Eq (9).  • From the graph, first peak, end peak, start time, and final time were noted • The average concentration was computed between the first and end peak as shown in Figure 9.
• With the calculated flow rate and the average concentration, the emission rate (E) was calculated using Eq (4).
• Using the emission rate (E) calculated in the previous step, the emission factor was calculated using Eq (8) and Eq (9).

PM2.5 (number) EFs:
The following are the steps involved in calculating the PM2.5 (number) EFs for each experiment: • First, the PM2.5 (small) readings were downloaded from the Dylos particulate matter analyzer and saved in an Excel spreadsheet • The sensor logs the values in # PM2.5 particles per 0.01 cubic feet of air for every 1 minute; the values were converted to # particles per cubic meter.
• After converting the values, a graph is plotted with time on X-axis and number concentration readings on Y-axis.
• Then, the same steps were used as discussed in calculating PM2. 5  The following are the steps involved in calculating the PM10 (number) EFs for each experiment: • First, the readings were downloaded from the Dylos particulate matter analyzer and saved in an Excel spreadsheet • The sensor logs the values in two channels; small (0.5µ<particles<2.5µ) and large (2.5µ<particles<10µ) in # particles per 0.01 cubic feet of air for every 1 minute; the PM10 number concentration was calculated by adding the small and large channels. The values were converted to # particles per cubic meter.
• After converting the values, a graph is plotted with time on X-axis and number concentration readings on Y-axis.
• Then, the same steps were used as discussed in calculating PM2.5 mass emission factors

Experimental Design
Experiments were conducted in four different cases (scenarios) to estimate the emission factors accurately. Table 1 shows the various cases used in this research.

Experimental Procedure
All the experiments involved in this research were conducted at the University of New Orleans.
The following procedure was followed for each experiment performed.
• Initial weight of the incense stick (before burning) was measured using a weighing balance • The whole box was cleared with ambient air, so the initial concentration is approximately zero, and the fans were turned on • The particulate matter and carbon monoxide sensors were initiated to log the readings This was to ensure the pollutant concentrations inside the chamber come to the initial levels.

Test Chamber
The test chamber required for this research is designed as per the EPA Method 1 guidelines. The setup with all the dimensions is shown in Figure 10. It is made of acrylic glass to see the smoke from the incense burning visually. The box was fabricated in such a way that all the openings were sealed to make it an airtight box. It has openings on one side to allow ambient air to enter as shown in Figure 10. It has a lid on the top that can be opened to setup the incense stick, a cup holder and a container as shown in Figure 11.
A fan was installed to draw the air through the rectangular duct as shown in Figure 11. The fan was fixed to a plate that can slide down as shown in Figure 11. Sensors were placed inside the chamber to record the pollutant readings. An incense stick in a cup holder was placed in a container with water; this was to ensure that the soot generated from incense burning does not affect the particulate matter readings. The actual laboratory setup of the test chamber is shown in Figure 12, 13.

Dylos Monitor
The Dylos DC 1700 particulate matter analyzer was used in the experiments as shown in Figure   14. It is a real-time analyzer that can be used for both ambient and indoor environments. It has an LCD screen that displays the number of particles in real-time; the left number represents the number of particles that are >0.5 microns and <2.5 microns in size, and the right number represents the number of particles that are >2.5 microns and <10 microns in size. It works on the principle of laser-based particle counter; it has an internal battery that runs for 6 hours of continuous operation. It has a real-time clock that stamps the date and time of each reading. Data can be logged in varying time intervals ranging from one minute to one hour. It has an internal memory that can store up to 10,000 data points, and the data can be downloaded by connecting the instrument to PC or laptop through RS 232 cable provided by the manufacturer [22] .

Industrial Scientific carbon monoxide monitor
The GasBadge Pro manufactured by Industrial Scientific was used to monitor carbon monoxide (CO) in the experiments as shown in Figure 15. It is a real-time CO analyzer that can detect values between 0 to 1500 ppm with 1ppm resolution. It is a portable instrument, runs on battery, and can be used in ambient and indoor air monitoring applications. The data can be logged in varying time intervals ranging from 2 to 300 seconds in 2-second increments. It has an inbuilt data logging capability that can store up to one year of data with one-minute logging intervals [23] .

Fan
An exhaust fan was installed to draw out air through the exhaust duct during the experiment. The speed of the fan can be controlled using a speed control unit as shown in Figure 16. The fan was run at 100% throttle for all the experiments.

Incense Sticks
Indian musk incense sticks were used for all the experiments as shown in Figure 17. These are hand rolled, manufactured in Mumbai, India. Incense sticks were placed in a cup holder inside the test chamber to perform experiments. The composition of the incense stick is not given by the manufacturer.

Weighing balance
A weighing balance was used to measure the weight of incense stick before (initial weight) and after burning (final weight) as shown in Figure 18. The readings measured were used in calculating the particulate matter and carbon monoxide emission factors.

TSI Air Velocity Meter
The Air Velocity Transducer 8455-09 was used in this research to estimate the air flow inside the exhaust duct. It is ideal in research and development labs for both temporary and permanent installations for air velocity measurement. It has a probe as shown in Figure 19; the probe is kept in the exhaust duct to measure the air velocity. The velocity reading (m/s) will be given on the screen as shown in Figure 20. The actual setup with a velocity meter and a stand to hold the meter is shown in Figure 21 [24] . The particulate matter and CO EFs from incense burning were reported in two sections, and they are as follows: 1. EFs based on flow rate computed using the velocity measurements and 2. EFs based on flow rate given by the fan manufacturer.
The EFs in each section is reported for four different cases/scenarios as discussed in the experimental design section. The EFs for different pollutants are as follows: • Carbon monoxide mass concentration

EFs based on flow rate computed using the velocity measurements
The velocity measurements were taken at 9 traverse points of the cross-sectional area of the exhaust duct as discussed in methodology section. The 9 velocity readings were averaged and used in the EF computations. The velocity measurement results and the average readings for one, two fans were presented in Table 2. The average velocity readings were then used to estimate the flow rate (Q) in the exhaust duct using the known cross-sectional area, the flow rate results were presented in Table 3.

Carbon Monoxide Emission Factors
The carbon monoxide emission factors from incense burning for four different cases is presented in Table 4. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 22) was presented to see the difference in EFs for four different cases.

PM2.5 (mass) Emission Factors
The PM2.5 (mass) emission factors from incense burning for four different cases is presented in Table 5. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 23) was presented to see the difference in EFs for four different cases.

PM2.5 (number) Emission Factors
The PM2.5 (number) emission factors from incense burning for four different cases is presented in Table 6. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 24) was presented to see the difference in EFs for four different cases.

PM10 (mass) Emission Factors
The PM10 (mass) emission factors from incense burning for four different cases is presented in Table 7. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 25) was presented to see the difference in EFs for four different cases.   (Fig 26) was presented to see the difference in EFs for four different cases. This is because of the Dylos sensor is not sensitive at higher particulate concentrations. The amount of smoke generated from two incense sticks was so high, the sensor was not able to measure those higher concentrations. Thus, the PM10 (mass and number) and PM2.5 (mass and number) EFs developed using two incense sticks were considered not reliable.

EFs based on flow rate computed using the fan manufacturer value
The air flow of the fan given by the manufacturer was 23 cubic feet/min for a single fan, theoretically for two fans, it will be 46 cubic feet/min. The air flow values were then converted to cubic meter/sec or liter/min and then used for computing EFs.

Carbon Monoxide Emission Factors
The carbon monoxide emission factors from incense burning for four different cases is presented in Table 9. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 27) was presented to see the difference in EFs for four different cases.  The PM2.5 (mass) emission factors from incense burning for four different cases is presented in Table 10. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 28) was presented to see the difference in EFs for four different cases.   Table 11. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 29) was presented to see the difference in EFs for four different cases.  The PM2.5 (mass) emission factors from incense burning for four different cases is presented in Table 12. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 30) was presented to see the difference in EFs for four different cases.  The PM10 (number) emission factors from incense burning for four different cases is presented in Table 13. Several test runs were performed for each case, and the average, standard deviation for the results were computed and reported in the table below. The bar graph (Fig 31) was presented to see the difference in EFs for four different cases. In this research, both (a) measured flow rates applicable for the experimental setup used and (b) the manufacturer supplied flow rates were used to compute EFs. The purpose was to see the variation in EFs using these two methods. However, it should be noted that the flow rates given by the manufacturer may be for unrestricted flow conditions. In our case, the flow was restricted and was only let into the test chamber through six openings of 1 inch in diameter. The measured flow rates reflect this fact as indicated by the reduced measured flow rate compared to the manufacturer supplied flow rate. Thus, the EFs computed using the actual/measured flow rates are more reliable.

Conclusions
Indoor air quality is a growing concern in the world; incense burning is a major threat to indoor  concentrations, so the EFs developed using two incense sticks (burning at the same time) were considered not reliable. The gravimetric method can yield reliable results, but it is cumbersome and expensive.
The EFs developed using the fan manufacturer air-flow value resulted in high EFs. This is due to the restricted air flow, due to a very few openings (six 1-inch holes on one side of the chamber) to allow the ambient air to flow into the chamber. Thus, the airflow indicated by the manufacturer might not be realistic for the experimental conditions used in this research.

Limitations and Future Recommendations
Experiments were performed using a single kind of incense stick because of time constraint.
Different kinds of incense sticks should be evaluated using this research. EFs for various incense sticks with different compositions must be evaluated and find the relationship between the composition of the material and pollutants emitted.
As the sensors (Dylos and Industrial Scientific CO sensor) were not based on Environmental

Appendix B -Experiment calculations EFs based on flow rate using velocity measurements
Carbon monoxide -Case 1, Experiment 1: