About This Episode
<On this episode of Conversations on the Creek, Zach May fills in as host to chat with Tim Greene of Verisk about wildfire risk and the role of smoke damage, and how advances in new data sources can help insurers better manage the risk.
The show kicks off with Tim recapping the past couple of years of wildfire trends.
From there, they delve more specifically into the role of smoke damage, including its financial impact during claims. Tim also shares his thoughts on what homeowners can do to protect themselves from smoke damage and how insurers can work more closely with their policyholders to mitigate potential losses.
Next, the conversation turns towards discussing new approaches to leveraging data to better assess smoke damage and utilizing it to make predictions of future losses.
Finally, the episode wraps up with Tim providing his own predictions of wildfire risk for the remainder of 2023.
Note: this episode was recorded prior to the 2023 Canadian wildfires, where smoke has been affecting air quality across a large portion of the U.S. Learn more about our partnership and integrations with Verisk by visiting .duckcreek.com/partner/verisk/.
Be sure to check out all our other episodes and follow us on Apple Podcasts, Spotify, and visit duckcreek.com/podcasts.
Transcript
00:13 Zach May:
From Duck Creek Technologies, it’s Conversations on the Creek, the podcast where we interview thought leaders about how the latest insurtech is transforming the P&C insurance industry. Whether you work in underwriting, sales and marketing, claims, or an insurer’s IT department, and each episode, we uncover the insights you need to create the future of insurance.
Hello, listeners, I’m Zack May. I’ll be filling in for Rob Savitsky as your host today. In today’s episode, we are so pumped to be joined by Tim Greene, a meteorologists and product and analytics manager from Verisk, who will share his perspectives on wildfire risk, and the role of smoke damage and how advances and new data sources can help insurers manage the risk.
Verisk is a leading strategic data analytics and technology partner to the global insurance industry. Hello, Tim. Welcome to the show. How are you today?
01:08 Tim Greene:
Hey Zach, I’m pretty good. Thanks for having me on the podcast today.
01:12 Zach:
We are excited to have you and excited to hear more about this, this fascinating subject and wildfires. So, with that, let’s just dive right in. So, Tim, it’s obviously been an active past couple of years with natural CAT events highlighted by things like hurricanes, wind damage, tornadoes in the south, you name it.
But today we are here, of course to talk about wildfires. So, could you recap the past couple of years and wildfires in the US and tell us what kind of trends have you been seeing?
1:44 Tim:
Yeah, absolutely. So, one of the things that we’ve seen at a broad scale is that managing wildfire risk is really something that’s grown quite rapidly in terms of a priority for insurers. And that’s largely driven by the uptick and frequency and severity that we’ve seen with respect to wildfires.
So, just looking at the last several years here, 2022 actually had more fires and more acres burned than 2021. So, 2022 had a little over 60,000 wildfires, and over 7 million acres that were burned. And it’s interesting too to see some of the emerging trends in where these fires are occurring. I think you and I are probably both naturally thinking about California as an area that sees a lot of wildfires.
But, actually last year, it was New Mexico that had led the pack with the Calf Canyon, Hermits Peak and Black Fires, each being over 325,000 acres a piece there. So, between the two of them, getting pretty close to a million acres out of that 7.2 million for all of 2022. But overall, we have seen just a growing trend here with wildfire risk going back even further. And this is largely driven by trends that we’ve seen in both climate and then exposure as well. And we’re going to start with that second one there, exposure.
So, the largest insured risk and losses are occurring in what’s known as the Wildland Urban Interface or WUI areas. And what that is, is it’s effectively where human development is meeting or co-mingling with undeveloped land. So overall, the wildlife urban interface is a little under 10% of the overall area of the continental US, but about a third of population and housing units are within that WIO.
So, I know speaking for myself, I live within the WUI in Central Massachusetts. What about you, Zach?
03:35 Zach:
Yeah, I am in Salt Lake City, Utah. So, I don’t know if I am technically in the WUI, but I’m going to guess I am or I’m at least very close to it.
03:44 Tim:
So there you go. 100% on the podcast here. So only 33% nationally, but 100% here today, so really just undergirding the fact that this is an emerging area of risk. And just to highlight some of the growth that we’ve seen in recent years, in the wildland urban interface there’s up to 4,000 acres of wildland that’s getting converted the wildland urban interface every day.
So, each year that’s about the equivalent of the state of Rhode Island that’s being added to overall wildland urban interface area. So, quite a lot. And then when you look at new home development, too, it’s about four out of every 10 new homes that are located in these new wildland urban interface areas.
So, you can almost think of wildfire as akin to hurricane, in the sense that growing exposure means that if the same wildfire were to happen today has happened 25, 30 years ago, it’d be a significantly higher loss, and certainly that’s something that we’ve seen with hurricanes that if the 1900 Galveston Hurricane were to happen today, it would be a much different scenario or same thing with some of the South Florida hurricanes even from as recently as the 90s.
04:55 Zach:
Sure, yeah. Fascinating, and coming back to the point you made of— in 2022 New Mexico being the state with the most acres burned in wildfires in the US. That is a surprising one for me. I just assume it’s California probably just about every year. So, really interesting.
So, Tim, when I think about wildfires, you know, naturally I’m— I’m always thinking of the fire itself. Right, you know, homes being burned to the ground, uprooting people’s lives. But let’s talk a little bit more of, maybe a hidden nuance of wildfires, and that is smoke and smoke damage. Where does smoke fit into wildfires and wildfire claims?
05:39 Tim:
Now, it’s a great point. And a great question, Zach, because wildfire claims are really a family of different types of loss. So obviously, like you say, being consumed entirely by flame is where you’re going to get your most catastrophic damage, but smoke, ash, soot, embers are also ways that you can cause property damage as a result of those wildfires. So again, not necessarily as catastrophic, but they can be very costly losses too.
Oftentimes, wildfire smoke claims have high rates of litigation associated with them. And that’s partially related to the lack of data availability that insurers have historically had related to wildfires. But it’s important too to kind of peel apart the onion here too, and understand what wildfire smoke is actually doing, what types of damage it can cause.
So, certainly, you can get the smells of— if you’ve ever been around wildfire area, you can really smell for a significant amount of time after the fact, some of the different chemicals, just soot, sometimes burning rubber-type smells that linger in the air. I already mentioned that I’m in Central Massachusetts, and I know from some of the wildfires up in British Columbia over the last several years, you can actually smell them as far to the east as where I’m located.
So, you can understand too how when you’re in a very concentrated area of wildfire smoke, you can get a lot of issues with smell, which is really just hazardous chemicals at the end of the day, that are seeping into different porous surfaces. But you can also stain wood, and plastic and fiberglass and porcelain. You can actually corrode metal. So with different surfaces, you can actually destroy or effectively reduce the structural integrity of certain things that would necessitate the entire replacement.
And it’s really important to understand that timing is a critical component of responding to wildfire smoke, and that there are certain surfaces that can be destroyed effectively or require replacement within just a few minutes of exposure. So that’s your upholstery, your porous materials or surfaces; Carpeting is an example.
But then there’s the hours to days as well, where things like walls and flooring and wooden furniture that maybe are suffering some staining, they can be cleaned if taken care of quickly. But if not quickly enough, then you might be looking at a complete replacement there. And then even longer than that, you can still have some ability to be able to, to clean things, but timing is really a critical component to the overall process.
08:08 Zach:
I never would have guessed smoke could be so harmful, and wildfires, you know, beyond just the fire burning itself. And it feels like you know, you’re noted that this can be a tricky area for things like litigation that comes from smoke claims. So really what I feel like I’m hearing you say is just some of the uncertainty about around being able to make more concise measurements on something like smoke can actually prove quite costly to insurers, really, because of that uncertainty.
08:40 Tim:
Yeah, absolutely. Absolutely.
08:43 Zach:
Do you have any other real-world examples of smoke or even ash causing damage to certain industries?
08:52 Tim:
Yeah, definitely. I think that’s a— that’s a great question, because we’re not just talking about homes and businesses and vehicles that are impacted by wildfires.
But one of the more interesting stories of the last several years in California, obviously being a wine country, is we’ve seen a lot of wildfire damage to wineries in that area, particularly in Napa, and Sonoma counties and in 2017 and 2020 specifically, where in each of those two years, over two dozen wineries were either completely destroyed or damaged in each of those, and as a result of between the two of them over $3 billion of total insured losses.
So, this includes things like the total damage or structural replacement. It included bottles that were being shattered by exposure to heat or fouled— the wine actually being fouled by that heat. And then vineyards of course, being burned, turned to scorched earth as well, where grape bushes are no more. That obviously can be an insured loss there too.
So, even though with some of those wineries and vineyards that were spared from the flames, they’re not necessarily out of the woods and this is what we observed, where there’s a condition known as smoke taint, which is actually where the smoke and ashy residue can permeate grape skin as it sits on the surface, and can tarnish the flavor of a grape and then have a resultant wine to the point where the grapes may no longer be viable as a source of wine, be it high end, or otherwise.
So, this leaves vintners with a very difficult decision to decide between potential reputational damage, between having a smoky-tasting wine product, or the economic hardship of having to pass up on a year’s worth of wine altogether.
10:33 Zach:
Fascinating. You know, I certainly don’t think I’d want to drink smoky grape wine, but I don’t know, maybe that’s the thing I’m missing out and don’t fully know about.
10:44 Tim:
Yeah, smokey is something that you think about maybe with your whiskey or your bourbon, but I don’t think too many people are looking for that in their wine.
10:50 Zach:
Sure, not me. So, just moving on. So, you know, you’ve already covered smoke to an extent, I want to— I do want to stay on that topic, because I find it quite interesting. And, again, bringing more attention to these areas of wildfires that might not always be front of mind. So, you know, say an insured policyholder submits a claim due to smoke damage, how are insurers actually assessing the smoke damage? And how are they servicing smoke damage claims today?
11:27 Tim:
Yeah, so a lot of the responses that we’ve heard from the insurance industry is that oftentimes there’s little information to base some of their analysis off of, other than knowing where wildfires are generally occurring from public databases. California Fire is a great resource in the state of California, but really a limited amount of information related to smoke itself.
So, particularly important here, because smoke is a highly localized peril, in that there are very small areas that are being exposed to dense amounts of smoke. And that can really vary based on the direction of the wind, the growth and location of the wildfire, then also, where in that wildfire perimeter, there’s an act of fire occurring at the time.
So, some of the things that we’ve heard over the years are like the air quality index, which you may have heard of. This is a yardstick that generally is measuring air pollution. And so typically, you’re seeing things like ground level ozone, carbon monoxide, and particulate matter.
And now recall that smoke is— since it was just soot and ash mixed together in an air aerosol, which is effectively just particulate matter. And so, you got to understand, though, that with particulate matter, you’re also looking at things like pollen and soil dust.
So, for a public health risk index, which is what the AQI is in nature, it does a fine job. But insurers really need more specific information, more curated information, as it pertains to that surface level smoke in order to better understand that hazard and deploy resources, both competently and effectively.
13:03 Zach:
So, keeping on the subject of smoke for a little while longer, you know, let’s— and in relating insurance and how claims are covered, let’s talk the money aspect. What are some of the financial impacts of smoke claims from wildfires that you are seeing and some of those trends?
13:22 Tim:
Yeah, so the actual amount of damage can really vary from fire to fire, and from claim to claim, and it depends on a number of factors. So, exposure is a big part of it. And I would suggest that maybe we think about it like the boulder fire that occurred at the tail end of 2021, and then at the very beginning of 2022. That was a very small fire, but it impacted a large number of structures and buildings over the short amount of time and the small area that it was affecting.
So, exposure is a critical component. How much wind is another component— is the smoke quickly being dispersed? Fuel type is another important one there too. So, are we talking about dense, very moist forests or dried out shrubs, or fine grasses that can affect the amount of smoke and the different particles that are suspended a lot as a result of the smoke?
But then I would say in general, what we’ve seen is on average between $10,000 and $15,000 per claim for wildfire smoke claims. So, important again to note that delays and high litigation rates can really drive those numbers up significantly there. So, another thing too, is that oftentimes we find that insurers are generalizing their wildfire types of losses, like smoke or soot, or ember damage, just as general— a wildfire related loss.
But, this can really make it difficult to flesh out some of the impacts of those ancillary loss types without additional data layers, like the geospatial data or imagery for both wildfires and smoke.
So, when you’re really breaking it down, the proper cause of loss classifications are essential for that accurate collection under reinsurance contracts, and misclassifying a claim can have very direct impacts on overall loss ratios. So, the more detail the better. And the more information that you can have on each individual wildfire related claim, whether it’s flame, smoke or ash, the more information that you have, the better.
15:20 Zach:
That makes that makes total sense, and I feel like aligns with all insurance claims, the more info, the more data, the more accurate the cause of losses, the better for all parties involved.
Yep. So, being good, I guess as a good policyholder and, and things that the insureds can be doing to— maybe protect themselves from things like wildfire and potentially even smoke, although I get that can be impossible to do sometimes, you know, to what extent can homeowners or you know, policyholders, property owners protect themselves from smoke damage, specifically? And how can insurers actually work with their policyholders to really help mitigate these potential future losses?
16:06 Tim:
Yeah, so with smoke specifically, it’s very critical to seal off every potential entry point to the home as soon as you have an understanding that there is going to be a high amount of wildfire smoke in your area. So, what necessarily does that entail?
That means powering off and sealing your HVAC system, so probably pulling out the old box fan and relying on that on the inside of your house, but keep your windows closed as you’re doing so. And also avoiding activities that could quickly deteriorate your indoor air quality, like using aerosol products, burning candles. So, think about that, if there’s a power outage too, or even vacuuming. And granted, you can do a pretty good job with some of the latest HEPA filters.
But even still, if you can avoid doing that, you’re going to be in pretty good shape. And then maybe the most important part for protecting your home against smoke damage is as soon as air quality improves outside, time to bring that air inside, because there’s a good chance that you have some pretty stale air that’s accumulated inside the home. And so, that’s when you can open the windows up, put the box fan and get that going and try to clear things out a little bit.
But then getting to the second half of your question about what insurers can do and how we can better interface with the policyholders. There is some new legislation that’s coming out of California, where insurers are facing new regulatory measures that have been intended to address the rating of property coverage in wildfire-prone areas.
So, the series of regulations is going to require them to account for and reflect some specific mitigation factors, and the rating plans that are filed with the California Department of Insurance. And the intention here is ultimately to promote greater transparency around wildfire mitigation efforts at both the community level and then that individual property level as well. So, it’s not just smoke that we’re talking about here, but it’s also just general wildfire prevention.
So, building a defensible perimeter and removing potentially combustible objects or surfaces away from the home. So, that could be potentially vegetation or ornamental plants. And important to note too, I mentioned that this is in California, that this is going to go into effect, but other states are watching this closely as well— Colorado and Oregon are a few that have been mentioned.
So, I think this is something that we’re going to start to see more of, as states trying to be more proactive and trying to get insurance companies to offer more benefits to policyholders for better defending their homes against not just wildfire smoke, again, but the overall peril of wildfire and ultimately saving everybody a lot of money in the process, hopefully.
18:42 Zach:
They’re as high up as state legislation. You know, things like insurance best practices, right, the education that they can promote to their— to their policyholders. But even the small little things that you mentioned a minute ago, you know, candles, lack of vacuuming, bringing fresh air into the home once that air is clean; It’s those little things sometimes and I know, I’m sure can make a huge difference.
19:08 Tim:
Yes, certainly. And then even at a macro level, you can look at it and approximately 85% of wildfires are manmade in nature, at least as far as an ignition source is concerned. So, definitely room for improvement. They’re not necessarily at an insurance level, but just widening the education there. Because oftentimes, it can be just pure negligence that leads to those situations.
19:31 Zach:
Of course, sadly, it is, often too many times. So, what’s turned the page here to newer ways in which insurers can perhaps better measure, assess, understand the financial impact of smoke and even other perils resulting from wildfires. What are some of the new approaches of utilizing that data to better assess, you know, this smoke damage? How is that data being gathered and maybe as important, is that data trusted and reliable?
20:04 Tim:
Yeah, so there have been a lot of technological advances that we’ve all been a party of over the last 10, 15, 20, 25 years. So, computer processing speeds have really shot through the roof. Cloud computing is something that over the last several years, has really opened the doors to a lot of new developments, specifically around new computer models, some even being capable of being run in real time of modeling things, lots of different perils, but wildfire and wildfire smoke being two of them.
So, a rising tide lifts all boats. And that’s what we’ve seen here with natural hazard models. So, not just talking about hail, wind, tornado and hurricane anymore. But wildfire and wildfire smoke models are gathering greater and greater spatial resolution, better temporal resolution, and as a result, greater accuracy there. So, getting back to how these models can be effective and how insurance carriers and adjusters are able to use this is to accomplish a few key tasks.
So, number one, you need to estimate the extent of surface level wildfire smoke, and then the intensity of smoke as well. Two important characteristics there, and then also accounting for low-level winds and then topography too; How that’s impacting where smoke is going to go, and how quickly it’s going to dissipate, then modeling the dispersion and deposition of smoke over space and time as well.
So, how much ash is going to be effectively rained out of a column of smoke, and how quickly is a column of smoke going to disperse over area, oftentimes, wind can play a key component in that too. And then lastly, being tailored to insurance use cases.
So, we talked about the air quality index before. And that’s more of a public health metric. And so, it really doesn’t have that strong background in insurance to be able to provide the information that carriers need to say, yes, this area was exposed to high levels of smoke for a long time.
And they were also, again, significantly high to be able to cause potential damage if smoke were to seep into the home. So, really, the accuracy is a result of some of the advanced computational methods, new satellite data that we were able to obtain as well. But overall, I think insurers are going to have to rely on something that is created with insurance in mind.
22:23 Zach:
And as a follow up to that, you know, beyond the viewing something that already has occurred. What about predictions on this kind of data?
22:34 Tim:
Yeah, so I think as is often the case, there’s a gap between what’s publicly available and what’s actionable here.
So, in order to kind of forecast out what’s going to happen in the future, I think you need to start with that real-time component and understand what’s happening to my book of business right now, before first notice of loss, and starting to get a sense for— I have this many policies that are seeing very elevated levels of smoke concentration.
So really, some of the common use cases I think we’re going to start to see more of, is just that understanding before first notice of loss, where those claims are going to come in from. Also starting to estimate some of those losses, ensuring that staffing levels in different areas are appropriate, even proactively reaching out to impacted policyholders.
That’s something that we’ve seen with some of the other perils, natural hazards as well with some of the new models that are coming out, creating automated claims workflows to make more straight-through claims processing. So, yes, this policy looks like it’s very high-smoke. Let’s pay it out. Let’s not worry about this one. Let’s focus our efforts more on adjusting this claim, which looks a little bit more questionable.
And then also issuing moratoria for some of the impacted areas, which can be really important for some of these more insurtech-savvy carriers that are allowing you to start an insurance policy on your phone. You don’t want a wildfire to be in place, and then already have a policy that’s coming in there. So, using these data to help set more efficient moratoria can be helpful as well.
24:06 Zach:
Okay, so we’ve read past a little bit of the past couple of years of trends, statistics around wildfires, but— and really talked about data sources and what might be emerging there, right? But now let’s— let’s look ahead, what are your predictions for the upcoming 2023 wildfire season? Is the snowy winter out in some parts of the west going to help us, or is it all irrelevant at this point, and we’re just going to have another crazy season?
24:38 Tim:
Yeah, you’ve definitely seen a lot of snow so far this— I shouldn’t say so far this winter, but over the course of the last winter out in Salt Lake City, and so not a very big surprise then, that thus far there’s been well below-average wildfire activity.
In fact, so far this year, acres burned are less than half of what we saw at this point in 2022. But of course, can change in a hurry. Significant drought conditions are still prevalent in the southern and central plains. And that’s likely to persist into the summer months here.
But to your point, some of the record-setting snowpack and snow water equivalents that we’ve seen, have left the basins in the Sierra Nevada, Utah and Arizona, as well just fully charged, and I think we’re probably going to see snow melt lingering on into July this year. In some cases in Utah, there were ski resorts that recorded over 700 inches of snowfall, and not to distract from the fact either that the broader region, or much of it anyway, so, over 500 inches of snowfall.
So, as you may expect then, that kept drought conditions pretty much at bay, in quite a different scenario as we enter peak wildfire season here than what we saw last year at this time. But about 95% of the West was experiencing either abnormally dry, or just downright drought conditions. But so, looking ahead, the expectation for this year is somewhat of a lethargic start to wildfire season as a result of all the moisture. So, we’ve seen thus far, a robust green-up that’s coming on the heels of that.
But then something else that’s a little bit more of an unknown at this point is the some uncertainty with the El Nino Southern Oscillation teleconnection or— and so if you’re familiar with that, and so this is a question of how quickly we’re going to go from ENSO neutral to El Nino conditions. And what El Nino means for wildfire is that generally going to see more wet conditions for southern portions of the US along an axis of what’s called the subtropical jet stream, and then warmer temperatures for some of the more northern portions of the West and the plains.
Another thing to note too, is that a higher-than-average grass crop, which was fed by a steady stream of snow melt does somewhat raise the specter of elevated potential for grass fires. Once that village vegetation cures over the course of the summer, although, there has been some recent history that shows that there’s a weak correlation between these wet winters and very active grass fire season; so absent strong level winds, these fires typically have a higher rate of quick containment.
And as for potential carryover fine fuels from last year in areas where maybe there was a lot of dry fuel that was on the ground, but it did not ignite. The weight of heavy snowpack from this past winter is presumably compacted that to the point where windborne propagation is expected to be minimal. So, again, overall expecting a reduced start to wildfire season this year, but there’s no reason that there can’t be significantly more activity later in the year.
We already talked about the boulder fire on this conversation, and that happened again the very end of December and into January. So wildfire is really a 365-day a year season at this point. So, there’s no reason why things can’t change dramatically in the latter half of 2023.
28:12 Zach:
Well, I for one, I’m going to remain always optimistic, or maybe the lack of a big wildfire season this year, but as always, very unpredictable. So Tim, again, thank you for sharing your knowledge and joining us today. I feel like you will continue to stay busy and really help weed this part of the industry forward.
28:35 Tim:
Yeah, no, it was great to be here talking with you today, Zach. So, thank you.
28:40 Zach:
Tim, thanks again, and thank you all for listening today. You can learn more about our partnership and integrations with Verisk by visiting duckcreek.com/partner/verisk. If you enjoyed this podcast, be sure to check out all of the other episodes and follow us on Apple podcasts, Spotify and duckcreek.com/podcasts. We’ll see you in the next episode.
Related Links
- Learn more about our partnership and integrations with Verisk by visiting duckcreek.com/partner/verisk/.
- Check out all our other episodes and follow us on Apple Podcasts, Spotify, and Google Podcasts.
Featured On This Podcast
![](https://www.duckcreek.com/wp-content/uploads/2023/11/95x109-Tim-Greene.png)
Tim Greene
Product Analytics Manager,
Verisk
![](https://www.duckcreek.com/wp-content/uploads/2023/10/Podcast_image_3.png)
Rob Savitsky
Director of Product Marketing,
Duck Creek Technologies
You May Also Be Interested In
In today’s episode, recorded following a panel discussion at the Insurtech NY 2024 Spring Conference, Andrew Schwartz, a Senior Analyst at Celent, joins host Rob Savitsky to explore the use of generative AI and advanced analytics in insurance claims management.
Given the rise of generative AI, how are today’s bots different from those of the prior generation? What are the biggest opportunities for insurers to leverage them across their operations?
We explore how critical payments is to the overall insurance value chain, why insurers struggle to comprehend how much they are even spending on the payments capabilities they provide, and discuss the potential savings insurers can extract from a next generation payments setup.