Research team: Caroline Frischmon, Barbara Weckesser, Dr. Kathy Duderstadt. Not pictured: Leah Burks and Jennifer Crosslin
In February and March of this year, we partnered with Caroline Frischmon and Dr. Kathy Duderstadt to conduct air monitoring for a total of 28 days in our neighborhood using low cost sensors, called HAQ-Pods developed by Hanningan Air Quality Lab. The sensors detected Volatile organic compounds (VOCs) and Ammonia. They were not able to detect all pollutants of concern, such as sulfur dioxide, hydrogen sulfide, and nitrogen oxides. We also pulled data from the existing purple air monitor located in the community to measure particulate matter to conduct our analysis. Additionally, seven residents also recorded symptom and odors over the course of the two months.
We placed two pods inside the neighborhood and one outside the neighborhood on the other side of town (as shown in the picture below.) With this data, we were able to determine the number of pollution episodes and duration of the episodes. We then took the episodic data and compared it with symptom and odor reports, industry self reported data, and wind direction.
Our findings
We recorded more nearly 3x more pollution episodes in our neighborhood than outside of it. 27 and 31 (from the two pods inside the neighborhood) compared to 11 outside of the neighborhood.
Reports of symptoms and odors often coincided with detection of episodes.
When we look at the days where all pods were working properly (indicated by the green dots), we find this to be true for all episodes detected except six. There were four days where residents reported symptoms and odors but not episodes were detected. One possible reason for the discrepancy is that we the sensors did not detect the pollutant that may be causing the smell or health symptoms. Additionally, human noses are incredibly sensitive. Sometimes more sensitive than technology. There were two days when episodes were detected but no odors or symptoms were reported. Reasons for the discrepancy in these instances include 1) participants don’t always go outside, so they could miss an episode; and 2) not every pollution plume reaches every house in the neighborhood. There were some differences between southwest and northeast Cherokee Forest.
The duration of the episodes were on average about 35 minutes. The longest recorded episode lasted 2 hours and 13 minutes and the shortest 2 minutes. This is important to note because many air monitoring studies are set up to take measurements over a 24 hour period. We noticed in our study that when we average over a 24 hour period, the pollution episodes disappear which could lead some to make claims that the air quality poses little to no risk to health of the residents.
The wind direction during episodes shows they are coming from industry.
Keeping symptom and odors are important for getting a fuller picture of residents' experiences and the potential health impacts of exposure to pollution. As previously mentioned more often than not, symptoms and odor reports coincided with episode detections, such as the four days following a Chevron flaring incident on February 7th.
"Three days straight for odor. Opened side door. Smell overwhelmed me."
Just as we were setting up our equipment, on February 7th Chevron reported a flaring violation, releasing sulfur dioxide. The following four days, we detected frequent episodes. The reports of symptoms and odors were so bad that the Jackson County Emergency Response and Chevron visited the neighborhood.
On the other hand, on February 18th and 19th, our pods did not detect any episodes even though residents reported strong acid smells. These reports were particularly significant because 4 of 7 households reported burning eyes, sore throats, headaches, and two people vomited, a third felt nauseous.
"I just got up from nap. feels like something is making air heavy. My eyes are on fire. I’m almost afraid to open door. I started feeling sick then nauseous and vomiting."
Comments