Reducing the Wildland Fire Threat to

Understanding how ignitions occur is critical for effectively mitigating home fire losses during wildland fires. The threat of life and property losses during wildland fires is a significant issue for Federal, State, and local agencies that have responsibilities involving homes within and adjacent to wildlands. Agencies have shifted attention to communities adjacent to wildlands through pre-suppression and suppression activities. Research for the Structure Ignition Assessment Model (SIAM) that includes modeling, experiments, and case studies indicates that effective residential fire loss mitigation must focus on the home and its immediate surroundings. This has significant implications far agency policy and specific activities such as hazard mapping and fuel management. The threat of life and property losses during wildland fires is a significant issue for Federal, State, and local fire and planning agencies who must consider residential development within and adjacent to wildlands. The 1995 USDA Forest Service Strategic Assessment of Fire Management (USDA Forest Service 1995) lists five principal fire management issues. One of those issues is the “loss of lives, property, and resources associated with fire in the wildland/urban interface” (p. 3). The report further identifies “the management of fire and fuels in the wildland/urban interface” as a topic for further assessment. Because this is more than a Forest Service issue, the National Wildland/Urban Interface Fire Protection Program, a multi-agency endeavor, has been established for over a decade and is sponsored by the Department of Interior land management agencies, the USDA Forest Service, the National Association of State Foresters, and the National Fire Protection Association. This program also has an advisory committee associated with the multi-agency National Wildfire Coordinating Group. These examples indicate that the wildland fire threat to homes significantly influences fire management policies and suggests that this issue has significant economic impacts through management activities, direct property losses, and associated tort claims. The wildland fire threat to homes is commonly termed the wildland-urban interface (W-UI) fire problem. This and similar terms (e.g., wildland-urban intermix) refer to an area or location where a wildland fire can potentially ignite Homes. A senior physicist at the Stanford Research Institute, C.P. Butler (1974), coined the term “urban-wildland interface” and described this fire problem: In its simplest terms, the fire interface is any point where the fuel feeding a wildfire changes from natural (wildland) fuel to man-made (urban) fuel. ...For this to happen, wildland fire must be close enough for its flying brands or flames to contact the flammable parts of the structure (p. 3). In his definition, Butler provides important references to the characteristics of this problem. He identifies homes (“urban”) as potential fuel and indicates that the distance between the wildland fire and the home (“close enough”) is an important factor for structure ignition. How close the fire is to a home relates to how much heat the structure will receive. ____________________ 1 An abbreviated version of this paper was presented at the Symposium on Fire Economics, Policy, and Planning: Bottom Lines, April 5-9, 1999, San Diego, California. 2 Research Physical Scientist, Fire Sciences Laboratory, Rocky Mountain Research Station, PD. Box 8089, Missoula, MT 59807. e-mail: Jcohen/[email protected] USDA Forest Service Gen.Tech.Rep. PSW-GTR-173. 1999. 189 Session IV Wildland Fire Threat to Homes—Cohen These two factors, the homes and fire proximity, represent the fuel and heat “sides” of the fire triangle, respectively. The fire triangle—fuel, heat, and oxygen— represents the critical factors for combustion. Fires burn and ignitions occur only if a sufficient supply of each factor is present. By characterizing the home as fuel and the heat from flames and firebrands, we can describe a home’s ignitability. An understanding of home ignitability provides a basis for reducing potential W-UI fire losses in a more effective and efficient manner than current approaches. Ignition and Fire Spread are a Local Process Fire spreads as a continually propagating process, not as a moving mass. Unlike a flash flood or an avalanche where a mass engulfs objects in its path, fire spreads because the locations along the path meet the requirements for combustion. For example, C.P. Butler (1974) provides an account from 1848 by Henry Lewis about pioneers being caught on the Great Plains during a fire: When the emigrants are surprised by a prairie fire, they mow down the grass on a patch of land large enough for the wagon, horse, etc., to stand on. They then pile up the grass and light it. The same wind, which is sweeping the original fire toward them, now drives the second fire away from them. Thus, although they are surrounded by a sea of flames, they are relatively safe. Where the grass is cut, the fire has no fuel and goes no further. In this way, experienced people may escape a terrible fate (p. 1-2). It is important to note that the complete success of this technique also relies on their wagons and other goods not igniting and burning from firebrands. This account describes a situation that has similarities with the W-UI fire problem. A wildland fire does not spread to homes unless the homes meet the fuel and heat requirements sufficient for ignition and continued combustion. In the prairie fire situation, sufficient fuel was removed (by their escape fire) adjacent to the wagons to prevent burning (and injury) and the wagons were ignition resistant enough to not ignite and burn from firebrands. Similarly, the flammables adjacent to a home can be managed with the home’s materials and design chosen to minimize potential firebrand ignitions. This can occur regardless of how intensely or fast spreading other fires are burning. Reducing W-UI fire losses must involve a reduction in the flammability of the home (fuel) in relation to its potential severe-case exposure from flames and firebrands (heat). The essential question remains as to how much reduction in flammables (e.g., how much vegetative fuel clearance) must be done relative to the home fuel characteristics to significantly reduce the potential home losses associated with wildland fires. Insights for Reducing Ignitions from Flames Recent research provides insights for determining the vegetation clearance required for reducing home ignitions. Structure ignition modeling, fire experiments, and W-UI fire case studies provide a consistent indication of the fuel and heat required for home ignitions. The Structure Ignition Assessment Model (SIAM) (Cohen 1995) assesses the potential ignitability of a structure related to the W-UI fire context. SIAM calculates the amount of heat transferred to a structure from a flame source on the basis of the flame characteristics and the flame distance from a structure. Then, given this thermal exposure, SIAM calculates the amount of time required for the occurrence of wood ignition and flaming (Tran and others 1992). On the basis of severe-case assumptions of flame radiation and exposure time, SIAM calculations indicate that large wildland flame fronts (e.g., forest crown fires) will not ignite wood surfaces (e.g., the typical variety of exterior wood walls) at distances greater than 40 meters (Cohen and Butler [In press]). For example, the incident radiant heat flux, the amount of radiant heat a wall would receive from flames, depends on its distance from the fire. That is, the rate of radiant energy 190 USDA Forest Service Gen.Tech.Rep. PSW-GTR-173. 1999. Wildland Fire Threat to Homes---Cohen Session IV per unit wall area decreases as the distance increases (fig. 1,). In addition, the time required for a wood wall to ignite depends on its distance from a flame front of the given height and width (fig. 1). But the flame’s burning time compared to the required ignition time is important. If at some distance the fire front produces a heat flux sufficient to ignite a wood wall, but the flaming duration is less than that required for ignition, then ignition will not occur. At a distance of 40 meters, the radiant heat flux is less than 20 kilowatts per square meter, which corresponds to a minimum ignition time of greater than 10 minutes (fig. 1). Crown fire experiments in forests and shrublands indicate that the burning duration of these large flames is on the order of 1 minute at a specific location. This is because these wildland fires depend on the rapid consumption of the fine dead and gure 1 IAM calcul nergy/unit the minim live vegetation (e.g., forest crown fires). Fi S ates the incident radiant heat flux (e -area/time reaching a surface) and um time for piloted ignition (ignition th a small ignition flame or spark) as a nction of distance for the given flame size. e flame is assumed to be a uniform, parallel ane, black body emitter. Experimental fire studies associated with the International Crown Fire odeling Experiment (Alexander and others 1998) generally concur with the IAM calculations. Data were obtained from instrumented wall sections that were laced 10 meters from the forest edge of the crown fire burn plots. Comparisons etween SIAM calculations and the observed heat flux data indicate that SIAM verestimates the amount of heat received. For example, the SIAM calculated pote iven nonflammable roofs, Stanford Research wi fu Th pl M S p b o ntial radiant heat flux for an experimental crown fire was 69 kW/ sq meter as compared to the measured maximum of 46 kW/sq meter. This is expected since SIAM assumes a uniform and constant heat source and flames are not uniform and constant. Thus, the SIAM calculations for an actual flame front represent a severecase estimate of the heat received and the potential for ignition. The SIAM distances represent an upper estimate of the separation required to prevent flame ignitions (fig. 1). Past fire case studies also generally concur with SIAM estimates and the crown fire observations. Analyses of southern California home losses done by the Stanford Research Institute for the 1961 Belair-Brentwood Fire (Howard and others 1973) and by the University of California, Berkeley, for the 1990 Painted Cave Fire (Foote and Gilless 1996) are consistent with SIAM estimates and the experimental crown fire data. G _______________________ 3 Unpublished data on file, Rocky Mountain Research Station, Fire Sciences Laboratory) Missoula, Montana. 4 Unpublished data on file, Rocky Mountain Research Station, Fire Sciences Laboratory, Missoula,