infrastructure · environment · family: the rulebook was written for a world that no longer exists
fire codes that protect against the wrong fire
Wildland-urban interface fire codes are largely unproven — building codes address the wrong fire exposure
Problem statement
Over 46 million U.S. residential structures are at risk from wildland-urban interface (WUI) fires, but NIST research has found that "WUI fire codes and standards remain largely unproven to actually mitigate WUI fire spread and structure ignition processes." Current building codes reference fire test methods (ASTM E119, UL 263) designed for urban fires — flame contact and radiant heat — but NIST's review of the literature reports that surveys of several WUI fires find the majority of structure fires are ignited by embers (firebrands), not direct flame. The fire exposure building codes test for is fundamentally different from the fire exposure that actually destroys WUI buildings.
Why this matters
WUI fires destroy thousands of structures annually, and the problem is worsening: NIST describes WUI fires as becoming more and more common and the economic toll of wildfires as steadily rising. Building owners who comply with existing fire codes may still lose their structures because the codes are not based on realistic WUI fire exposure science. Traditional fire codes address individual buildings, but WUI fire is a community-level phenomenon — one burning structure generates firebrands that ignite neighbors — requiring community-scale risk assessment, which NIST proposed to add to codes through a community fire-exposure zoning system.
What’s been tried and why it hasn’t worked
NIST developed the Fire Risk Reduction in Communities Program and a proposed WUI Hazard Scale. They built specialized firebrand generators for laboratory research and conducted post-fire data collection on parcel vulnerabilities. A NIST researcher convened ISO TC92/WG14 ("Large Outdoor Fires and the Built Environment"), whose charge was to develop the technical report ISO TR/24188. However, "WUI fire science is much less developed than more mature areas of fire safety science." The physics of firebrand generation, transport over long distances, and ignition of building assemblies are unsettled: NIST says firebrand generation from structure combustion "remains unsolved, and no rigorous modeling attempts have been made." NIST's literature review finds that "available empirical data on the ember ignition of solid wood substrates are limited and is not sufficient to predict the ignitability and sustained smoldering of wood products exposed to an arbitrary set of conditions they may face in the WUI." Community-level fire modeling requires integrating vegetation, topography, weather, and structure vulnerability, and NIST says science-based engineering tools for rigorous evaluation of fire spread in WUI fuels "remain elusive." Even if NIST develops the science, adoption into the ICC's International Wildland-Urban Interface Code and then by communities is a further step: NIST notes that improvements to codes and standards "must be adopted and enforced to fully realize the benefits."
What would unlock progress
Validated fire exposure standards based on firebrand flux rather than flame contact — essentially, testing building assemblies against the actual threat. This requires quantifying firebrand generation rates from burning vegetation and structures, and developing standardized ember exposure test methods that building materials can be certified against. A community-scale vulnerability model that accounts for structure-to-structure fire spread (the cascade effect) would enable risk-informed code requirements at the neighborhood level rather than building-by-building.
Entry points for student teams
Building a firebrand generator is a real door but not a facility-free one: even NIST's reduced-scale continuous-feed design needs a burn room with a wind supply, exhaust, and the fire-safety infrastructure around it, so this arm requires a fire protection engineering department that already operates a certified burn facility and a faculty sponsor who can put student work on it. The door that needs no facility is an ignition-pathway audit against what codes actually certify: CAL FIRE's Damage Inspection (DINS) records are published openly on California's open-data portal (https://data.ca.gov/dataset/cal-fire-damage-inspection-dins-data) with per-structure construction-feature fields — roofing, siding, eaves, vent screening, deck and window attributes are in the DINS data dictionary — alongside a damage class, so a team can test which assembly details separate destroyed from surviving structures inside the same fire perimeter and then name the specific ember exposures that no listed test method (ASTM E2886 ember intrusion, ASTM E108 roofing, ASTM E119 flame contact) covers. A second facility-free door is the GIS neighborhood spread model, built from LANDFIRE fuel and vegetation layers (https://landfire.gov/), USGS 3DEP terrain (https://www.usgs.gov/3d-elevation-program), and county assessor parcel footprints, producing structure-to-structure risk maps at the block scale. Relevant disciplines: fire protection engineering, materials science, GIS/geospatial analysis, mechanical engineering, statistics.
Genes
Structural cousins: stuck the same way, in other fields
Sources
NIST WUI Building and Fire Codes and Standards Project; NIST Fire Risk Reduction in Communities Program (the quoted sentences are on its project page "Reduced Ignition of Building Components in Wildland-Urban Interface (WUI) Fires," ); Nazare, Leventon and Davis, NIST Technical Note 2153, "Ignitibility of Structural Wood Products Exposed to Embers During Wildland Fires: A Review of Literature," April 2021, go to source 1 ↗ go to source 2 ↗ go to source 3 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
- The `failure:wrong-problem` tag is strongly warranted: fire codes test for the wrong fire exposure. This is not a case of missing standards but of existing standards addressing the wrong physics.
- The insurance industry's retreat from WUI areas removes a market-based incentive for improved standards — a vicious cycle.
- The 2025 LA fires make this acutely timely; political will for code reform is at a peak.
- NIST's firebrand generator designs are publicly available, making student replication feasible.
- ISO TC92/WG14 work may provide an international pathway that circumvents slow U.S. code adoption.
- Note reconciled 2026-08-20: the first note above argues for `failure:wrong-problem`; in the August taxonomy pass that tag was replaced by `failure:proxy-metric` — the goal (structures that survive WUI fire) was always right, but the certified quantity (ASTM E119/UL 263 flame/radiant exposure) diverges from the actual ignition mechanism (embers). The original note is kept verbatim as the tagging rationale of record.
Reconciliation 2026-08-21: Entry-point realism (C37 triage, score 1) — flag CONFIRMED. "Build a small-scale firebrand generator (NIST has published designs)" is not student-scale on its own: even the reduced "Baby Dragon" needs a burn room with a wind supply, exhaust, and fire-safety infrastructure, so that arm was kept but given an explicit access line (a fire protection engineering department operating a certified burn facility, plus a faculty sponsor) rather than deleted. Unflagged-door finding: the GIS spread model, which triage assumed would carry the brief, is genuinely facility-free but its "publicly available parcel data, vegetation maps, and topography" were unnamed, so the door was made concrete on resources verified by fetch today — LANDFIRE fuel/vegetation layers (https://landfire.gov/, "free of charge" full-extent downloads) and USGS 3DEP terrain (https://www.usgs.gov/3d-elevation-program, "available free of charge and without use restrictions") — with parcel footprints left as county assessor data, whose availability genuinely varies by county. Added a third, fully facility-free door that did not exist before: an ignition-pathway audit using CAL FIRE's Damage Inspection (DINS) dataset, verified open on California's portal under CC-BY with CSV/shapefile/GeoJSON downloads and per-structure construction-feature and damage-class fields (https://data.ca.gov/dataset/cal-fire-damage-inspection-dins-data, fetched), read against the test methods codes actually list (ASTM E2886, E108, E119) — this puts the brief's core claim within reach of a team with no fire lab. One earlier Source Note above states "NIST's firebrand generator designs are publicly available, making student replication feasible"; the designs are indeed public, but replication is facility-gated, and the entry-points section now says so — the original note is kept verbatim as the record. Declined to cite: the CAL FIRE Office of the State Fire Marshal Building Materials Listing (osfm.fire.ca.gov returned 403 to fetch), and no NIST post-fire parcel dataset URL is named because none was verified in this session.
Reconciliation 2026-09-30: Round 6 read of the whole write-up against its sources, all fetched 2026-09-30 as full text. The Source-line page (https://www.nist.gov/programs-projects/wui-building-and-fire-codes-and-standards-project, which now redirects to .../wildland-urban-interface-wui-building-and-fire-codes-and-standards) and the Fire Risk Reduction in Communities Program page (https://www.nist.gov/programs-projects/fire-risk-reduction-communities-program) were read in full, as were the program's linked WUI project pages, and NIST TN 2153 was read in full as a PDF from https://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.2153.pdf. Two of the three quoted sentences are not on the Source-line page or the program page; NIST's site search located them on the program's "Reduced Ignition of Building Components in Wildland-Urban Interface (WUI) Fires" project page (https://www.nist.gov/programs-projects/reduced-ignition-building-components-wildland-urban-interface-wui-fires), which also carries the ISO TC92/WG14 facts; that page is now named on the Source line as the location of the quotes. Changes: (1) The Source line's short title for TN 2153 was replaced with the title, authors and date copied from the document itself ("Ignitibility of Structural Wood Products Exposed to Embers During Wildland Fires: A Review of Literature," Nazare, Leventon and Davis, April 2021, https://doi.org/10.6028/NIST.TN.2153). (2) The first quotation ended at "structure ignition"; the page reads "structure ignition processes," and the quotation now does too. (3) "NIST demonstrated that the majority of WUI structure ignitions are caused by firebrands" overstated the source: TN 2153 is a literature review and says "it is evident from the surveys of several WUI fires that the majority of structural fires are ignited within the structures by embers," citing another review; the sentence now says that. (4) "The 2025 Los Angeles fires caused hundreds of billions in damages" and "development pushes into fire-prone landscapes and climate change extends fire seasons" are in no source and were removed; the trend sentence now uses the program's own words (WUI fires "becoming more and more common," on the Reduced Ignition page; "the economic toll of wildfires has been steadily rising," on the program page, so the sentence says wildfires, not WUI fires, for the toll). "Thousands of structures annually" was kept; the program's "Improving WUI Community Fire Protection" project page says "Since 2000 over 38,000 homes have been lost to WUI fires," and the January 2025 Wayback capture of the program's parcel-vulnerabilities project page said "Between 2002 and 2016, an average of over 3,000 structures per year were lost to WUI fires in the U.S." (5) "Insurance companies are withdrawing from WUI areas entirely because there is no reliable way to measure or certify building resilience" is in no source and was removed. (6) "Community-scale risk assessment that no code framework provides" was softened to what the Source-line page says: NIST proposed adding a community fire-exposure zoning system to codes. (7) "NIST convened ISO TC92/WG14 and co-developed ISO TR/24188" now follows the page: a NIST researcher (Manzello) was approved as convener of WG14, whose charge was to develop the technical report ISO TR/24188. (8) "Transport over kilometer-scale distances" and "far more complex than flame spread" are not in any source; the sentence now says transport over long distances (the program's wording) and quotes the page's "remains unsolved, and no rigorous modeling attempts have been made" about firebrand generation from structures. (9) The quotation "currently little quantifiable information that links the ember generation from wildland fuels to building assemblies testing" could not be found on any NIST page fetched, in NIST's site-wide search, or in TN 2153; it was replaced with TN 2153's own sentence that available empirical data on ember ignition of solid wood "are limited and is not sufficient to predict the ignitability and sustained smoldering of wood products exposed to an arbitrary set of conditions they may face in the WUI." (10) "Scales that exceed current computational capability" is not in any source; the WUI Fire Spread and Modeling project page says science-based engineering tools for evaluating fire spread in WUI fuels "remain elusive," and the sentence now says that, without tying "elusive" to scale, which the page does not do (the January 2025 capture of the parcel-vulnerabilities page in fact calls NIST's Fire Dynamics Simulator "well-suited for studying structures and surrounding combustibles" at 0.1 m to 10 m grid resolution). (11) "Adds years of delay" is not in any source; the sentence now uses the ICC code's name as the Source-line page gives it (International Wildland-Urban Interface Code) and the parcel-vulnerabilities page's statement that code improvements "must be adopted and enforced to fully realize the benefits." (12) The nickname "Baby Dragon" is in no source; the entry point now says "reduced-scale continuous-feed design," the description TN 2153 gives for NIST's device. Confirmed in the sources and left as written: over 46 million residential structures at risk (Source-line page); the quotation "WUI fire science is much less developed than more mature areas of fire safety science"; ASTM E119 and UL 263 as fire-resistance standards that evaluate flaming and radiant heat exposures, and ASTM E2886 (embers, vents), ASTM E108 (roofing) in TN 2153's Table 1; NIST's WUI Fire Hazard Scale; firebrand generators built for laboratory research; post-fire data collection on parcel vulnerabilities; the ICC International Wildland-Urban Interface Code. The DINS dataset (https://data.ca.gov/dataset/cal-fire-damage-inspection-dins-data) was confirmed open under Creative Commons Attribution with CSV, Shapefile and GeoJSON downloads, and its feature service lists DAMAGE, ROOFCONSTRUCTION, EAVES, VENTSCREEN, EXTERIORSIDING, WINDOWPANE and deck fields; the LANDFIRE and USGS 3DEP pages open. The titles were not changed. The earlier Source Notes that mention insurance withdrawal and the 2025 Los Angeles fires are working notes kept verbatim; nothing in the named sources supports them. Genome tags untouched. A second reader went over all twelve repairs the same day, refetching every named source (the Source-line page and its 2017 and 2019 captures, the program page and its 2016, 2019 and 2022 captures, the Reduced Ignition page and its 2019 capture, the parcel, firebrand, spread and community-protection project pages, TN 2153 as PDF, the DINS record and feature service, LANDFIRE and 3DEP); confirmed the missing quotation in (9) is in none of them, in NIST's site search, or in a web search; kept ten repairs as written, tightened the wording of (4) and (10) as described above, and corrected the page attribution for the structure-loss figures in (4).