The field of civil engineering has long been involved in mitigating hazards affecting communities. Fires, however, have generally fallen outside of the purview of civil engineering, possibly due to the fact that civil engineering education does not provide the necessary background in fire science. Nonetheless, “resilience” is a term that spans the civil engineering and fire safety disciplines. In this blog post, I draw parallels between the hazards traditionally considered by civil engineers (e.g., earthquakes and hurricanes) and wildland-urban interface (WUI) fires with the intent of steering the fire safety engineering community toward a common language and approach. We will explore the facets of resilience and the tools and methods used by civil engineers to ensure resilience. 

What Causes a Disaster?

Natural hazards become disasters when a large number of people are killed/injured or the damage to infrastructure disrupts a significant portion of society’s activities. Hurricane Katrina (shown in Fig. 1), for example, was undeniably a disaster. 1,800 people died and a large portion of the city (estimated 80 percent) suffered extensive flooding due to levee failures. An estimated 1 million people were displaced by the storm and subsequent flooding, and the estimated damage was $125 billion dollars.

So what led to Hurricane Katrina being such a major disaster? The hurricane was downgraded to a Category 3 by the time the hurricane made landfall. While Katrina was a large storm, it does not even show up on a top ten list of most powerful hurricanes to hit the U.S. In actuality, most of the damage caused by Hurricane Katrina was due to the flooding that resulted from the failure of a number of levees and floodwalls. The failure was an engineering failure, in which the levee system was inadequately designed and constructed. There was an added layer of tragedy because the communities impacted most by the flooding were low-income. Hurricane Katrina contrasts with the Loma Prieta Earthquake of 1989 (shown in Fig. 2), for example, where buildings that were designed for seismic hazards performed well [1]. While 27,000 buildings were damaged in the earthquake, only 63 people died. However, the Loma Prieta Earthquake showed a flaw in the design methodology that did not account for the functionality of the buildings following an earthquake event.

Another natural hazard that was considered a disaster was the Joplin Tornado (shown in Fig. 3). The Joplin Tornado was an EF-5 tornado (high intensity) that hit the urban area of Joplin, MO. The storm resulted in 158 deaths and 1,150 injuries, and it caused $2.8 billion dollars in property loss. There were two factors that led to the Joplin Tornado being a disaster: (1) because the tornado was an EF-5, it was characterized by extreme winds that were capable of severe damage (i.e., the hazard was severe), and (2) it hit a city that was densely populated (i.e., an urban area).

Let’s now look at a wildfire disaster. The Camp Fire, which is shown in Fig. 4, started on November 8, 2018 and continued for over two weeks, burning more than 150,000 acres. It is known as the deadliest and most destructive fire in California’s history. It presented a major problem when the fire reached the town of Paradise. 85 people died and 17 were injured. The fire destroyed approximately 18,000 structures and caused an estimated $17 billion in damage. The intensity of the fire was aided by dry conditions, heavy grass cover, and high winds. Furthermore, few of the homes were designed to be resistant to wildfire.

In comparison to the three hazards we have considered so far, there are some important similarities. First off, the hazard was severe and it hit an urban area, resulting in massive losses. The disaster wasn’t necessarily the result of an engineering failure, but, arguably, it highlights a flaw in the engineering design process because so many homes were destroyed. It begs the question, What can engineers do to strengthen communities at the WUI? 

Design Practices

In civil engineering, design practices are either prescriptive or performance-based. Prescriptive design is rooted in standard testing procedures of building components. The result is a design process that is relatively simplistic and can easily be regulated. Prescriptive design tends to be overly conservative due to the simplistic approach taken. Performance-based design (PBD) on the other hand is based on an advanced understanding of the hazard and the resistance provided by the structure. Different performance levels (e.g., fully operational, operational, life-safe, collapse prevention) are chosen for different hazard levels (e.g., frequent, occasional, rare, very rare). PBD is based on testing that better represents reality, hence doing away with a lot of the conservatism built into prescriptive design. Models are much more advanced, and the design process tends to be a lot more complicated, leading to some challenges in getting designs approved. By quantifying the reliability of systems, engineers can use PBD to better optimize structures, resulting in significant cost-savings and increased sustainability. 

Looking at the history of codes and standards for seismic, wind, and fire hazards reveals some fundamental differences in how communities are designed for the respective hazards. On the one hand, the modern seismic design of buildings goes back over one hundred years to the 1906 San Francisco Earthquake, which caused great damage prompting extensive research in the decades that followed. As engineers’ knowledge of the seismic performance of structures improved, the building codes and design standards were improved. When the Loma Prieta Earthquake hit in 1989, the buildings largely met life-safety standards. The main shortfall, however, was that thousands of buildings could not be used [1]. The design philosophy shifted to include “functionality” as a performance requirement. This was the beginning of PBD. 

While earthquake engineering focuses largely on multi-storey steel and concrete structures, the damage produced by hurricanes is concentrated primarily on non-engineered dwellings. Because the structures are not engineered, design modifications are largely prescriptive. Any changes to the design are implemented in local building codes to make structures more resilient. In recent years, there has been an interest in adopting a performance-based design approach in wind engineering, but the interest is mostly to save costs on overdesigned buildings.

Wildfires are similar to hurricanes in that the structures affected are largely non-engineered dwellings. Unlike wind and earthquake events, which generate large forces on structures, wildfires pose a problem of structural ignition. Thus, resistance to wildfires can be achieved by limiting the fire threat (e.g., reducing the fuel in forests through controlled burning, creating defensible space between the wildland and the community) and hardening structures (e.g., using fire resistant cladding and roofing). Like hurricanes, design practices are built into local buildings codes, and changes are often made following notable wildfire events. While much of the earlier work on structural ignition under wildfires focused on ignition due to radiation from the fire, recent work has highlighted ignition due to firebrand showers as a key mechanism for structural ignition. The science is still evolving, but progress has been made to develop standard testing methods to determine the potential for structural ignition under firebrand showers [2]. 

Earthquake engineering is the most mature of the fields discussed here, with PBD readily incorporated in the current design process in earthquake prone regions. The research is mature, and the tools to analyze structural systems are well developed. Wind engineering has largely emphasized a prescriptive approach, but interest has developed in recent years to explore PBD to save costs on overdesigned structures. Wildfires, on the other hand, have only recently received attention from engineers. Most of the prior focus was on forest fire prevention, which was problematic because it resulted in an excess of fuel in most forests in North America. The excess fuel in combination with extreme weather conditions (e.g., drought, high winds) that support wildfire growth have created a major problem at the wildland-urban interface. Undoubtedly, the research will follow, and advanced methods of design may be applied in the future. Nonetheless, engineering for wildfires is relatively primitive compared to earthquakes and wind hazards, as illustrated in Table 1.

	Prescriptive Design	Performance-based Design
Earthquake	Advanced	Advanced
Wind	Advanced	In Progress
Wildfire	In Progress

Definitions of Resilience

A component of hazard mitigation is resilience. Resilience has various meanings depending on the field in which it is referenced. In ecological and socioecological systems, for example, resilience is a property of a system, whereas in psychology, resilience might refer to recovery from trauma [3]. In engineering, resilience is often a property ascribed to infrastructure systems, and it refers to the system’s ability to tolerate a shock and return to normal operations quickly. A review by Bozza et al. [4] found the definition of the resilience of urban systems varies according to the type of system and its properties. Systems that make up the civil infrastructure systems include the infrastructure systems, the safety management systems, the organizational systems, the social-ecological systems, the economic systems, the social systems, and the communities. 

Let’s see how the wildfire problem might be cast in two existing resilience models. First, for infrastructure systems, one definition of resilience involves overcoming negative consequences of a disaster and returning to normal operations as quickly as possible. In terms of overcoming negative consequences of a disaster, we can look to fire safety systems, which include forestry and land management; building design, construction, and maintenance; policy and its enforcement; fire service funding and preparation, communication systems, and egress planning and execution. These systems are interdependent and complex, but they share the common objective of helping a community overcome the negative consequences of a disaster. When we cast the wildfire problem within this framework, we can observe that success in meeting the objective is limited. One reason is that funds are limited–we don’t have infinite resources to put towards firefighting, land management, hardening of structures, and so on. Additionally, we run into the issue that policy lags behind the science. So while we know that creating defensible space can reduce the wildfire hazard, a community may not have policies in place to designate areas as defensible space. An additional challenge is associated with the practicality/cost of hardening existing structures. Lastly, a systems-level view of a community’s resilience to wildfire may call for new approaches to analyze and design the complex, interrelated system. 

In the same vein, we can study a community’s ability to return to normal operations as quickly as possible. If the negative consequences of the disaster are overcome, presumably the damage is limited and the return to operations will be swift. In communities that are not resilient, the rebuilding process can take a lot of time. There is a silver lining in cases like Paradise, CA: the damage from the wildfire was so extensive that community is left to rebuild nearly the entire city, and planners and developers can redesign the city to be more resistant in the event that a severe wildfire impacts the community again. Next 10 [5] explored three strategies for rebuilding Paradise following the Camp Fire:

• Analyze existing recovery plans with historical growth trends to guide the planned development (i.e., status quo). 
• Offer incentives to disaster survivors to move to lower risk locations while promoting infill development in existing urban nodes (i.e., retreating). 
• Rebuild some housing in high-risk areas while incorporating robust wildfire mitigation features like development clusters surrounded by defensible space (i.e., clustering)

Their analysis showed that retreating and clustering both reduced future fire risk. However, should one of these strategies be adopted over the status quo? To answer this question, we have to think about how to balance social and economic factors. If there is a housing shortage in urban areas that are safe from wildfire, for example, these strategies are not feasible. Similarly, people may not relocate if the incentives are insufficient. 

Let’s look at another resilience model that is commonly used in engineering: Robustness, Redundancy, Resourcefulness, and Rapidity (the Four R’s). These terms are defined as follows [1]:

• Robustness: the physical strength of components such as infrastructure
• Redundancy: the substitutability of components
• Resourcefulness: the ability to mobilize resources
• Rapidity: the ability to return to the pre-disturbance state in a timely manner

Table 2 gives some examples of system components that support each of the Four R’s. Since wildfires are different from seismic and wind hazards in that human activity can actually affect the severity of the wildfire hazard, I’ve extended the definition of robustness to include hardening of structures as well as human activity to remove fuel, create defensible spaces, etc. We don’t rely so much on redundancy to prevent structural ignition, but we do need redundancy in other fire safety systems, like egress and firefighting. In terms of resourcefulness and rapidity, a lot of the success of fire safety systems at the WUI depend on adequate funding of the fire service as well as financial planning and policy-making to protect the community and ensure it returns to normal operations as soon as possible. In other words, communities at the WUI should be prepared for the wildfire hazard. After a WUI fire, assessments should be done to, for example, see if the resources provided were sufficient and that the time to re-home the community was not excessive. Note that this list is not exhaustive but is merely meant to cast the wildfire problem in the resilience framework.

	Before	After
Robustness	Harden structures and remove fuel in advance of fire threat	Adopt a rebuilding strategy that reduces future threats
Redundancy	Egress planning, firefighter training/planning	Analyze unanticipated gaps and bottlenecks and correct them
Resourcefulness	Ensure fire service is adequately funded; financial planning and policy-making that anticipates the most extreme scenarios	Analyze shortfalls in funding/resources and their impact on the response
Rapidity	Adequate fire safety plan and execution	Assess the time to re-home the community

Conclusions

In this post, we looked at several natural hazards to explore the components of the hazards to led to disasters where loss of life was high and/or community operations were severely disrupted. We found that disasters were often linked to severe hazards occurring in densely populated regions and/or when engineering failures occur or the engineering process is flawed. By comparing wildfires to seismic engineering and wind engineering, we see that wildfire engineering is an immature field in terms of our understanding of the hazard and the codes and standards used to protect buildings. We then applied two resilience models to the wildfire problem. We can see that the models fit reasonably well, and by using the definitions given here we may be able to better engineer resilient communities at wildland-urban interface. The hope is that the wildfire community can take some of these concepts and further the engineering approach used to make communities at the WUI more resilient.

References

[1] P.F. Fratessa (1994). “Buildings,” in: Practical Lessons from the Loma Prieta Earthquake, National Academy Press. 

[2] Manzello, S. , Zhou, K. and Suzuki, S. (2014). “Experimental Study of Firebrand Transport,” Fire Technology, https://doi.org/10.1007/s10694-014-0411-8, accessed 9/22/22.

[3] Mochizuki et al. (2018). “An overdue alignment of risk and resilience? A conceptual contribution to community resilience,” Disasters, 42, 361-391.

[4] Bozza, A., Asprone, D., and Fabbrocino, F. (2017). “Urban Resilience: A Civil Engineering Perspective,” Sustainability, 9, 103. 

[5] Next 10 (2021). Rebuilding for a Resilient Recovery: Planning in California’s Wildland Urban Interface, https://www.next10.org/publications/rebuilding-resilient, accessed 9/22/22.