Wildfires that devastate mountain communities have the potential to foul the water distribution system running underneath residential structures. But knowing which water pipes have been affected is challenging. Erica Fischer, assistant professor of structural engineering, is working with a team of engineers and scientists to develop and test sensors that can easily indicate if water pipes need to be replaced following a fire.
[VIDEO CLIP: Wildfire Sound Installation, Jon Bellona YouTube, Sep 27, 2019]
[VIDEO CLIP: Sights and sounds from Camp Fire in Paradise, Sacramento Bee YouTube, Nov 9, 2018]
PALMER: On Thursday, November 8, 2018, a faulty electric transmission line triggered a fire near Camp Creek Road in Butte County, California. The fire, later dubbed the Camp Fire, quickly spread west threatening several mountain communities, including the towns of Paradise and Concow, on its way to becoming the deadliest and most destructive wildfire in California’s history. The fire destroyed more than 18,000 structures, mostly residential homes. Images of these smoldering structures, along with maps of road closures and power outages, plastered our screens for weeks. But what about the damage done to the network of underground pipes that shuttle water to and from residential homes?
I’m your host Chris Palmer. In this episode of Engineering Out Loud, we’ll focus on a crucial secondary impact of devastating wildfires: the potential for water contamination within the water distribution system. There are many possible reasons why this might happen. Through research, engineers and scientists can determine the causes of this contamination and figure out how to mitigate it so communities can rebuild with the assurance that their water is safe to drink.
[MUSIC: “The Ether Bunny,” by Eyes Closed Audio, licensed under CC by 3.0]
PALMER: From the College of Engineering at Oregon State University, this is “Engineering Out Loud.”
FISCHER: I've been going back and forth to Paradise, California, quite a bit. I've been out after the Eagle Creek fire, and then after last summer, I was out on route 126, where the Holiday Farm Fire impacted the neighborhoods of Blue River and Vida. So, it's complete devastation, is really what you see. When we have an earthquake or hurricane or a flood there's portions of the community that are impacted. And then there's portions of the community that are not impacted. With wildfires, what we tend to see is this mass devastation, it's kind of like a tornado.
PALMER: That’s Erica Fischer. She’s an assistant professor in the School of Civil and Construction Engineering, specializing in structures. Her research focuses on the resilience and robustness of infrastructure — that is, how buildings and lifelines perform in natural hazards, such as earthquakes, fire, and wind.
FISCHER: As I started getting into the wildfire discussions being led by people in the forest service and forest colleges who are so knowledgeable about wildfires. And we need them for wildfire research and for our wildfire work and mitigation, however, they're not civil engineers, they don't fully understand how communities function. And, so, I just think civil engineers need to be more involved in this conversation.
We understand how communities operate and we understand how institutions operate. Very often it's not the physical structure that really matters. You know, a lot of hospitals can operate without a physical structure. They can't operate without utilities though.
PALMER: As you might have guessed, one of the most important utilities in a community is its water system.
FISCHER: Basically, the way the water systems work is you have your main transmission line. That's your big pipe. So, when you're driving down the highway and they're working on building the water pipes and you see these massive sections, that's not really what's impacted by the wildfires. Those are underneath roads. You need something to burn for the fire to be intense, and there isn't much fuel on a road. So, we don't normally see a lot of damage to these main transmission lines.
PALMER: Between the main transmission lines running under roads and your home are pipes called service laterals. These lines can be made of many different types of materials such as steel, iron, or plastic. This is where Paradise, and another town, Santa Rosa, which was hit by the Tubbs fire, found the majority of water contamination.
[MUSIC: “Skeptic” by Podington Bear, soundofpicture.com]
FISCHER: Plastic pipes are very popular. They're very durable. They're great for corrosion resistance. They're great for seismic. They are inexpensive. They're installing them all over the place in Corvallis. Sometimes we'll have copper pipes. Copper's great, but expensive.
PALMER: There’s also older pipes made of ductile iron or steel or galvanized steel.
FISCHER: Between the road and your house, there is a lot of fuel. Like, think about your garden, think about your home. During a wildfire, all of that is burning and those pipes aren't buried very deep. So, all of this stuff is burning and it's burning for quite a long time. Homes burn for multiple hours. And, so, it's heating up the ground below it, it's heating up the pipes below it. So, a few things could happen. First, these pipes can heat up and contaminants can migrate from the pipe itself into the water distribution system.
Another thing that could happen is when your home is burning and completely burns down to the ground, the place where your pipe attaches to your house is now open. And what happens is because a lot of water is being used in the system because we're fighting fires, right, and they're attaching to the fire hydrants, this negative pressure builds up and ash and soot and everything that just burned in the air gets sucked into that opening.
PALMER: All that stuff can contaminate the water in the distribution system. Plus, imagine a smoldering chunk of charcoal being sucked back into those pipes.
FISCHER: It can heat up the pipe from the inside out, and then cause the pipe material to start migrating into the water. So, we don't really know what the one cause is of contamination. I'm also just talking about pipes here, but there's so many other components when you're attaching pipes to the water meter, the pipes to each other, there's these gaskets and there's valves and there's connectors. And we don't know how any of these behave at temperatures outside of the operational temperature of the pipe.
[MUSIC: “Skeptic” by Podington Bear, soundofpicture.com]
PALMER: To get a better sense of what happens to water pipes near scorched homes, Erica has made multiple visits to Paradise and other communities recovering from devastating wildfires in the past few years. Part of that recovery that she’s witnessed up close that could significantly impact residents’ long-term health involves figuring out whether pipes delivering water to their homes need to be replaced. Erica and colleagues from around the country are looking into a method for efficiently and systematically determining if the pipe damage is enough to cause health issues.
FISCHER: In both Paradise in Santa Rosa, they were sampling water all over the place. There was no methodology or system to do it. And they didn't know where the source of the contamination was. And, so, they ended up replacing all of the service laterals in the town, which is incredibly expensive and time consuming and disruptive to a community. First, we have to understand why this happens. Is there one sole contributor, or is it kind of this perfect storm of events? We’re seeing that particularly in Paradise, California, that they're installing backflow devices on your water meter. So, if there is negative pressure build up, whatever comes into the pipe from the water meter to the house can't get past the water meter and back into the main transmission line and then be circulated around the town.
We did see that in Santa Rosa that even places, even homes that had backflow devices there was contamination. So, is the water meter itself contributing to the contamination? We've seen photos of these water meters that are plastic and just completely melted. So, we have to be able to identify what is that contributor? And then we can come up with a way to mitigate for it.
PALMER: Erica is part of a team that is developing and testing a sensor that can indicate if the temperature inside a pipe has exceeded a safe limit.
FISCHER: The sensor, being developed by Professor Lauren Linderman at the University of Minnesota will indicate when the critical temperature of a pipeline has been exceeded.
PALMER: Rounding out the team are Lisa Ellsworth, assistant professor of fisheries and wildlife at Oregon State, Jenna Tilt, assistant professor of social sciences at Oregon State, and Brad Wham, assistant professor of infrastructure at the University of Colorado, Boulder.
FISCHER: The project is funded by the Alfred P. Sloan Foundation. They really wanted to target early to mid-career academics, which all of us at the time we applied were untenured. So, it was a really great opportunity for all of us. They really wanted to promote diversity and inclusion and four out of five of our PIs are women. So, we hit that mark pretty easily. We have really interesting conversations because we're all approaching this in just a totally different way. So, it really is a great collaboration of expertise.
So, we're doing tests here at Oregon State University to determine what that critical temperature is, so that Professor Linderman can tune it in to be specific to the pipes. It's actually a kind of cool sensor. It's a complete passive sensor. So, that means, it’s, there's no battery.
PALMER: Passive sensors emit radio-frequency identification (or RFID) signals that can be read with a handheld device. The team is developing sensors to be installed in pipes near homes at risk of wildfires. These communities are called wildland urban interface communities. After a fire, the sensors would be scanned to determine if temperatures inside the pipe exceeded a certain safety threshold.
FISCHER: You know, if the pipe is completely melted, obviously it needs to be replaced, and obviously there's going to be some issues. And where the sensor can be really helpful is, well, what happens when the pipe isn't melted? It doesn't have visual damage. Is there actual damage? Are contaminants migrating into the water at that point?
A lot of these WUI communities; wildland urban interface communities, are remote communities with small water districts. To burden these water districts with digging up pipes constantly and replacing sensors, that can become really expensive. So, we want, our goal was really from the get-go of finding something that you install it once, you forget about it, it's there and it can help you in the aftermath.
PALMER: Erica’s team has been testing different types of pipes by heating water inside them to different temperatures and then having the water tested at Edge Analytical Labs in Corvallis. They then have been using that data to construct numerical models of pipes in different environmental settings.
[MUSIC: “Elephants on Parade” by Podington Bear, soundofpicture.com]
FISCHER: So, we were able through those analyses to determine, you know, what kind of surface conditions or burial depths would actually allow us to get to those high temperatures that we think we would see issues with the pipes.
Simultaneously, we've been testing pipes to determine when do we see issues with the pipes? So, we've done this on a larger scale, and now we're kind of scaling down to do a number of experiments in a small furnace that we have here on campus, to just test a series of these pipes and gaskets and valves, just heating them up to elevated temperatures, keeping them there for a set amount of time, and then soaking them in water for about a week. And then testing that water again.
FISCHER: There are so many contaminants that we could potentially focus on, but what we are really honing in on are called BTEX compounds. So, benzene, toluene, ethylbenzene, and xylene. These are indicators of a greater problem.
PALMER: So far, heating up three-foot-long pipes filled with water up to 200 degrees Celsius has shown the water isn't exceeding the operational temperature of the pipe. And not surprisingly, they haven’t seen any contamination in the water either.
FISCHER: That in itself is, is a really big finding that we're able to heat this pipe up for multiple hours and the water inside the pipe isn't exceeding the operational temperature of the pipe. We are seeing damage to the pipe on the outside. We do see that the plastic is softening on the outside of the pipe. But it was surprising to us what a large thermal gradient we were seeing.
Now we're looking at heating the whole pipe, even the inside and small samples of the pipe, so that we do more of what's called a steady state test where we just heat the whole thing to one temperature. We've been really exploring how to even test these pipes and then using the data from our experiments to inform our numerical models. And then based on the results from our experiments, be like, ‘Yes, this situation would create contamination, or this situation would not create contamination.’ And when you can do that numerically, we can run tens of thousands of analyses and then we can start performing more of a probabilistic analysis and inform the community about what's the probability that you would see contamination in your pipe given these varying parameters.
PALMER: The long-term plan is to deploy these sensors in communities, where they will be able to withstand harsh environmental conditions, including fires. The information from the sensors can then inform communities after a fire where high temperatures were experienced in service laterals. However, Erica points out that the point of the sensor data is not to tell communities what types of pipe materials to build with.
FISCHER: Well, we don't particularly want to tell communities what materials to use and what not to use. That's really up to the communities. A lot of times it comes down to financial constraints and, again, plastic pipes are great for so many things, right. In general, it's more that communities should know where their high-risk areas are and what materials of pipes are everywhere.
So, we have a lot of sensor data already, like mapping data that we use post wildfires. Our goal is that this sensor can be integrated with the information we already use. And then it just empowers towns to have data to understand what happened. So, they can say, we don't need to replace the service laterals over here. It's going to be over here. Or where is the source of the contamination, where are the most heavily impacted areas? It also helps put this objective lens on responders. There's a lot of data and a lot of research showing that low-income, disenfranchised, vulnerable communities actually report damage less. They're not the squeaky wheels after the disaster. Yet, they probably have most of the damage or a lot more damage than some of the more wealthy, well-off communities. So, how can we focus our response and get clean water back online, quicker, in a more equitable way, is also what we're trying to do.
[MUSIC: “Deep Pools” by Podington Bear, soundofpicture.com]
PALMER: This episode was produced by me, Chris Palmer, with lots of help from Will Havnaer and the whole Engineering Out Loud podcast team.
Our intro music is “The Ether Bunny” by Eyes Closed Audio on SoundCloud, used with permission of a Creative Commons attribution license. The music and effects in this episode were also used with appropriate licenses. For more information, visit engineeringoutloud.oregonstate.edu.
FISCHER: I really do think that the work that I'm working on has been shaped by Oregon State University and the state of Oregon and the opportunities here. And it's a large reason as to why I even came here. I've watched Oregon over the years rely on Oregon State University and their scientists to inform policy. And it's happened on the earthquake front. It's happened on the tsunami front, and it's just so cool, you know, to live in a state that really uses their resources. It's really exciting as an engineer to see that and be involved in that.
