Air Quality, Quick and Dirty

8/30/19 Podcast

Air pollution degrades everything from cultural heritage sites like the Taj Mahal, to solar energy production. Duke civil and environmental engineering professor Mike Bergin is trying to fill in some of the knowledge gaps about what the pollution is and where it comes from, to better manage its effects.

podcast cover art: smokestack
Air Quality, Quick and Dirty


Michael Howard Bergin Profile Photo
Michael Howard Bergin Profile Photo

Michael Howard Bergin

Sternberg Family Professor of Civil & Environmental Engineering

Research Interests

Aerosols are generated by a variety of natural and anthropogenic sources. I am interested in the processes that emit, transform and deposit particulate matter. A…


M. Volborth: This is Rate of Change, a podcast from Duke Engineering dedicated to the ingenious ways that engineers are solving society’s toughest problems. I’m Miranda Volborth.

Mike Bergin: When you’re in an airplane and you look down and let’s say it’s a clear day, you can see the ground pretty well. You see, I don’t know, in the rural areas you see farms, trees, the country. Flying over a city, you see the city, buildings, stuff like that. But sometimes when the plane gets way up high, you’re above the clouds and it just looks white. And the reason that looks white is all those cloud droplets are scattering the sun’s energy back at us.

Mike Bergin: So when you get off a plane in Delhi, it’s very surprising. It looks like it’s super cloudy and overcast, but it’s not. It’s just all this whiteness from the intense amount of pollution. You definitely can’t see the sky often, and I would say the other thing about it is it seems dirty everywhere. There’s kind of a layer, a film of grit everywhere. And that grit is pollution. It’s on the buildings, it’s on the cars. If you go to your hotel building, it’s on the windows and if the windows have been open, it’s in the room and everywhere. So I think the particles are so small that little particles of pollution get everywhere.

M. Volborth: Mike Bergin is a professor of civil and environmental engineering at Duke University, though you probably wouldn’t be surprised if he told you he was a bike mechanic or that he had spent the last 30 years touring with some obscure indie rock band. He’s easy to laugh and he expresses a certain distrust for “the man,” as it were.

M. Volborth: But in reality, Mike has spent his entire adult life trying to understand air pollution and to educate people about it. In recent years, he’s forged strong bonds with collaborators in India, particularly at the Indian Institute of Technology at Gandhinagar, and the group is neck deep in research seeking to better understand the effects that air pollution has on human health and on the economy.

M. Volborth: I’ve never been to Delhi, or any of the places Mike goes to conduct his research, and I’ve only really experienced poor air quality during events like forest fires. If you haven’t either, this might put it into perspective.

Mike Bergin: Durham is a really clean place. Durham has a particulate matter level of roughly five to eight micrograms of particulate matter per cubic meter of air. And that’s one of the cleaner places in the US, really one of the cleaner urban areas. The World Health Organization’s standard is 10, so we consider 10 and below really clean. Delhi would be something like maybe one to three hundred micrograms per cubic meter. So 10 times above the standard, sometimes 1,000 or 2,000. So the levels are just off the charts, for sure.

Mike Bergin: Many years ago, probably 10 years ago now, maybe a little bit more, I was invited to a conference, and it was in Agra where the Taj Mahal is. We went on a tour of the Taj Mahal, and I noticed that they had scaffolding up around a big part of it, and part of it looked pretty white where the scaffolding was, and part of it looked really brown. I was asking them about it and they said they were cleaning it off, and it’s really, really a hassle to clean off. It takes a long time, a lot of effort.

Mike Bergin: I asked him why it was and they said it has something to do with pollution. And so I thought, wow, what does that do with pollution? And then they told me a bunch of things that didn’t really make sense to me. You know, they said, “Oh, there’s a big plant over there emitting a bunch of junk that’s depositing to the Taj Mahal.”

Mike Bergin: Since I know a little bit about it, I knew that one plant close by was not going to have that big of an influence. And I also knew that the city itself was, as I said before, India is a really hazy place and I knew the city was super hazy, so I knew that stuff was wafting in and probably depositing on the Taj Mahal.

Mike Bergin: So to make a long story short, got a little bit of money, me and my collaborators in India, and we did a little project there and basically we got little marble pieces. The Taj Mahal’s made of marble, and we cleaned them and we stuck them on the Taj Mahal. They let us do that. Then we came back like three, four months later and we collected and we looked at what was on them. And sure enough, it was pollution and it was kind of soot, and it was soot and particulate matter from burning stuff.

Mike Bergin: The way people dispose of their trash in India is typically they put it outside their houses right outside the front door. Then usually what happens is that gets gone through by various folks outside of their houses who take things that are of use, and leave the rest. Then at some point in time, and it looks to me like it occurs on a weekly basis, whole neighborhoods get together and they get their trash put in a big pile and they burn it. It goes right up into the air. So I think consequently a large fraction of the trash made in India go straight up into the air.

M. Volborth: Breathing that air has consequences. There are profoundly negative cardiovascular, respiratory, and neurological effects from exposure to pollution at that level.

M. Volborth: But to begin to manage air quality, you should first have an emissions inventory, a database detailing the type and amount of pollutants regularly discharged into the atmosphere. Without it, any regulatory actions could be called arbitrary. But data on residential trash burning for whatever the reason hasn’t really been included in emissions inventories. Maybe the relatively small fires seem nominal in the grand scheme of things, or maybe that data is just too difficult to capture.

Heidi Vreeland: Plumes are pretty difficult to measure, especially outdoor plumes of smoke, because the wind can blow them all kinds of directions. They change size. It is hard if you have a sensor that’s only at one point, like a lot of our monitoring historically has been from stationary monitors. Like the monitor’s on the ground, maybe 50 feet away the air quality’s totally different.

M. Volborth: Heidi Vreeland is a PhD candidate in Mike’s lab, and was one of the leads on a collaborative project with Duke, IIT, and also RTI International that sought to develop new approaches to monitoring air quality to help fill in some of this critical data. The group tested their approach in the neighborhoods around Ahmedabad.

Heidi Vreeland: The sites we were going to were just like these small residential alleys. It was very easy to find trash burning everywhere. This is very, very, very common. These are small-scale piles that range from, I don’t know, maybe a couple shoes’ size to, I don’t know, like a car maybe in area of piles that were constantly just smoldering. You drive by tons of smoldering piles all the time.

Heidi Vreeland: Something that is also important for residential and roadside burning, like I mentioned before, it’s hard to measure partially because the spatial area that that covers is very large and the signals are maybe small, but together they’re probably a big deal. That’s what most people think. So we were interested in trying to figure out how to cover a larger ground to look more at how to measure this large spatial area. And that’s actually where the idea for drones came in.

Heidi Vreeland: We have a low-cost sensor that we designed at Duke, and we attached that to the bottom of a drone, just a commercial drone. So having something that’s portable is pretty revolutionary, and we’re getting better at having more low-cost sensors and being able to deploy larger sensor networks so that we can use mathematical models to estimate the spatial area of what is the air pollution looking like over the city.

Heidi Vreeland: But in the past, we had to measure everything with maybe satellites or a balloon. We’d just send it up with a sensor, or it’d be a stationary monitoring site. I guess we’ve also used planes, but all these things are very expensive. This high cost has been a big limiting factor in the past.

M. Volborth: Do people see the drones and get excited and know what it is, and like-

Heidi Vreeland: Oh yes. And that is why our original plan and protocol was we decided to change things a little bit because drones attracted a good bit of attention. A lot of spectators came out to look at the drone. So that could have changed behavior or… We’re measuring particulate matter, an air pollution metric, so just little particles. So that also includes if people are stomping around outside, they’re kicking up some dust and things, and that might get measured. So we didn’t want anyone also standing really close to trash fires, which is where the drone was. So we’re like, “We should do a different direction.”

Heidi Vreeland: So what we decided to do after that was, we also were planning to do this, but look at a big municipal dump site. It’s a 84-acre dump site in Ahmedabad, has 80-foot walls of trash, and is just constantly smoldering. It’s huge. So we’re interested in looking at the air quality there. And just thinking about its sheer size, it’s clear how a drone could be much more useful in measuring the air quality over that big area than me with a sensor or a stationary monitor. So-

M. Volborth: Not to mention the safety.

Heidi Vreeland: Oh yeah, the safety.

M. Volborth: Like you can’t get in the middle of an 84-acre-

Heidi Vreeland: Absolutely. There was glass everywhere. Everything was on fire. Yeah. The students were not about to climb 80 feet of trash to go collect some data. Yeah. So that’s been a big thing with drones. Drones are also used to measure wildfires and volcanic plumes. Things that, yeah, people can’t access because they would be dangerous.

M. Volborth: I asked Heidi what the group found, and if she was surprised by anything that they learned.

Heidi Vreeland: There’s a lot more plastic in India than there probably is here. Most of the waste that we saw was plastic.

Heidi Vreeland: I was just thinking about how we actually collected trash from these different sites, and then we measured them in the field, but also collected samples to burn in a more controlled environment to try to get an idea of the characterization of these different emissions. So that required us sorting through a lot of trash, encountering some scary things like diapers. But yeah, so.

M. Volborth: And that’s the kind of stuff that just gets burned. I mean, everything just gets burned.

Heidi Vreeland: Right.

M. Volborth: Diapers, plastic.

Heidi Vreeland: Right. Oh yeah. We saw car seats, all kinds of stuff. So yeah, some things burn better than others, I’ll say that.

M. Volborth: The bottom line is that all this plastic trash when burned is sent straight up into the air. And those particles of burnt matter appear to be a big contributor to India’s air pollution problem.

Mike Bergin: They’re not good for the environment. They’re depositing in the environment, they’re getting transported around the environment. The different chemical species, people are breathing them in. In India, there’s maybe two to five million people dying per year because of breathing these particles in. So we can educate people about that at the same time, and then we can say, there’s also a huge benefit for renewable energy if you can decrease these pollutant concentrations. We’re hoping that we’ll figure out ways to connect with the public and show them this information in a way that would maybe make folks in the government spring to action.

Mike Bergin: So, MAPSOLE stands for minimizing air pollution’s impact on solar energy production, and it’s a project that we’re beginning here at Duke with a bunch of our collaborators. The idea is is that currently we know very little about air pollution’s overall impact on solar energy production. So we’ve brought together a team of people from industry around the world, some national labs from various places around the world, as well as a bunch of our university collaborators.

Mike Bergin: As it turns out, this problem is pretty complex to solve. So on one hand we’re trying to develop remote-sensing techniques using satellites to look down and see if solar panels are dirty. That requires people that know about remote sensing, requires people that know about data analytics and image processing. Also we’re trying to develop sensors that we can place at these places. Really cool, low-cost, autonomous, cheap sensors that can sense how dirty panels are getting that we put on these solar farms.

Mike Bergin: But we’re also really excited because we think we’re developing tools that can be used at even the rooftop solar level so every day a satellite can look down on every location where there’s a solar panel, as long as we can see through the clouds, and it can give you information about how your home solar panels are working or give your information if you’re in the Arabian peninsula about how your solar farm’s working, or your large solar farm in India or China.

Mike Bergin: We’re also developing really sophisticated tools that allow you to predict into the future what’s going to happen with your solar panels, how dirty they’re going to get, when you might want to clean them, when it would be best. For example, you wouldn’t want to clean them if it’s really going to rain really hard.

M. Volborth: Michael Valerino is another PhD student in Mike Bergin’s lab, and he’s working on this aspect of the MAPSOLE project figuring out when solar farms should clean their panels. It sounds like a small thing, but it’s really complex, and reliable database recommendations could have huge economic impact.

M. Valerino:  Basically you have to have people on either modified tractors or modified trucks with big brushes with water that go along and clean rows of panels. You imagine if you have a field with tens of thousands, hundreds of thousands of panels, how long that could take and how costly that good could end up being.

M. Valerino:  Right now, we estimate that this is somewhere between a $10 and $50 billion a year problem globally. So that has to do with lost energy as well as extra cleaning costs. So I mean, as you could imagine, especially in places without access to constant supply of clean water like India and the Middle East, places like that, drier climates, the cleaning costs can be pretty significant. So the overarching theme of the project is trying to get a better understanding of global soiling to allow energy producers to make more informed decisions regarding soiling mitigation.

M. Volborth: For example, common sense would tell you that if it’s been very dry and dusty, you’d want to clean your panels more frequently, right? But Michael says that humidity makes soiling worse because the particles get sticky and adhere to the panels. The tail end of the monsoon season actually saw soiling rates twice as high as normal. Effects like these that are seasonal and cyclical could shift the schedule by which solar farmers clean, and save them tons of money.

M. Valerino:  So our lab definitely focuses on the low cost. So as part of this project we’ve developed a couple of low-cost tools. The one I could talk about is the LAMPS station. We like our acronyms. LAMPS stands for low-cost alternative to Monitoring Photovoltaic Soiling. Basically it uses two small five-watt, about the size of an 8″ x 10″ sheet of paper, solar panels next to each other, and a custom software and hardware, which on the hardware side, thank goodness for computer engineering undergrads because that is not my background, but they were very helpful in designing a lot of the circuitry and whatnot.

New Speaker:  

M. Valerino:  So this LAMPS tool that we’ve developed can monitor soiling for about $200 a system. Right now we’re comparing it to kind of our reference soiling station, which is on the order of $5,000 to $10,000 a system. And it seems to be working within a few percent.

M. Valerino:  As you could imagine, a lot of the lack of information in the soiling field comes from most studies being single location. Of course they’re single location if each station is $10,000, right? But now with a $200 tool, you can start to get a better idea of global soiling simply because you can have more places that you’re monitoring soiling.

M. Volborth: It’s kind of baffling, just the discrepancy in price. Like why is what’s available commercially so expensive? That’s crazy.

M. Valerino:  It is. It is crazy. I got this question at a conference once, and I was presenting a poster and somebody said, “Well, why is your station so much cheaper than this other one?” And I wasn’t really sure. I mean, to be honest, it is kind of baffling.

M. Valerino:  So in terms of our station and a lot of what we do, I think it’s almost you could think of it is that diminishing… What’s the economic… Diminishing returns, right? So to make something 5% more accurate, you’re going to have to spend twice as much money to make it another 5% accurate. You’re going to have to spend-

M. Volborth: Never heard that.

M. Valerino:  … twice as much money, right? It’s kind of like your returns are diminishing for your effort or money.

Mike Bergin: I think often solving important problems does not take sophisticated instrumentation. I think there’s a time and a place for that, and I would say that, for example, there’s, in my field, environmental science and engineering and particularly in environmental monitoring, what tends to happen is as time goes on and people make more and more sophisticated instruments to look at more and more of the small nuances of particular pollutants, and I think that kind of research is really important. But then at the same time, when you think about it, just measuring something real simple like the total amount of particulate matter, the total amount of dust in the air in India, that’s not a rocket science problem, but there’s very few measurements of that all around India because there’s just not the available equipment.

Mike Bergin: So we’ve kind of focused on trying to get measurements to people so they can get an idea of what their environment is like around them, whether the environmental pollutant levels are fast and dirty, and so that they can understand what it means and then go and do something about it. So I think that’s kind of been our approach, and we do some more sophisticated measurements as well, but increasingly we’re going to try and do real simple stuff.

M. Volborth: If you found this research interesting, do keep your eyes on the Bergin Group: Mike has begun to look at the relationship between air pollution and mental health, and is investigating ways to min social media for health data as it relates to air pollution. Stay tuned.

M. Volborth: Thanks for listening. Subscribe for updates from Duke Engineering, and if you learned something from this podcast, please share it with others.

Sound effects from

Keep Listening

Explore all seasons of the Rate of Change podcast, dedicated to the ingenious ways engineers are solving society’s toughest problems.