Expert review in progress — published October 28, 2026

Research report

Resilient Structures

The strongest residential structure for the hazards on a county, and the runner-up. Colour is the top system on performance. Coastal high-wind and surge counties use the thin-finish exterior score. Search a county or ZIP, then read the hazard mix, the top three scores and the evidence labels.

Written by: BuiltStrong.ai research
Published: 2026-10-28
Funding: Independent. No manufacturer reviewed or funded it.

The map

How the scoring works

BuiltStrong ranks structures first. By default the map colours each region by the system with the strongest demonstrated performance against that region’s hazards. Price and permitting are scored too, but they count only if you move their sliders.

Three axes, each 0 to 10

Best structure measures how the whole house performs, walls, roof and the roof-to-wall connection together, against wind and windborne debris, tornado missiles, roof uplift, earthquake, fire, embers, flood, termites (including hidden paths through foam to wood), trapped and latent moisture, decay and corrosion, and what fire and debris do to the exterior face of the walls. It is scored on demonstrated performance and engineering: test results, ratings and load-path mechanics. 10 means strong performance; 1 means weak performance or a failed test. Cost, labour, the length of the code path and the amount of evidence do not enter it.
Best potential price is a rough ordering of likely build cost in the region: 10 is the cheapest likely option. It uses the cited cost studies, census framing shares and whether local plants and crews exist. No neutral, current, regional cost study exists, so treat it as low confidence.
Ease of permitting scores the code path: 9–10 for prescriptive IRC sections, about 7 for evaluation-report systems inside their limits, about 5 where an engineer must design it under ACI 318, and 2–3 for alternative-material approval with special inspection or a system outside its IRC limits.

Evidence badge: shown beside each score, never inside it

A score says how well a system performs. The badge says how much independent data stands behind that number. It appears beside each score in the click panel and in the capability table below, and it changes no total.

Extensive data A code-prescriptive path plus independent field or lab evidence (FEMA assessment teams, IBHS, Texas Tech missile tests, NFIP TB-2).
Moderate data An evaluation report, a product approval or an independent lab test (ICC-ES reports, the Florida approval, furnace ratings).
Emerging data A manufacturer-commissioned test, a single study, or engineering reasoning with no test yet (for example the SCIP tornado missile, or any price score).
Engineered / limited formal testing Scored on an engineering basis: a steel-reinforced concrete section designed by rational analysis and sealed by a registered engineer, with no published test of that specific assembly (the SCIP roof slab for uplift, roof impact and roof fire). Counted as no formal test in the averaged badge; it does not lower the score.

For a region’s structure score the badge is the weight-averaged depth of the hazard rows behind it; when that average is thin and engineering-basis rows outweigh the emerging ones, it reads “Engineered / limited formal testing”. Price is always “emerging”. Permitting is “extensive” where a prescriptive IRC section sets the path, “moderate” where an evaluation report does, and “emerging” where only alternative-material approval applies. The previous edition added 0.5 for the strongest evidence and took 1 off the weakest; that adjustment is gone.

Weighting

The weighted total is (structure × its slider + price × its slider + permitting × its slider) ÷ the sum of the sliders, so it stays on the 0–10 scale. The sliders are relative: 50/50/0 gives the same result as 100/100/0. With every slider at 0 the map falls back to structure only. The higher total wins. A tie-break, on the score for the region’s most heavily weighted hazard, applies only when two totals are equal after rounding to the hundredth. The panel marks that case with an asterisk. A gap of 0.2 is not a tie.

How the structure score is built

Structure score = the sum over hazards of (region hazard weight × system capability). Capability comes from the report’s Section 5.2 capability matrix, anchored on the dots (strong = 9, middle = 6, weak = 3, none or failed = 1) and adjusted for the performance the tests and ratings show. Evidence depth (Section 2.1) does not move the number.

The whole house, not just the wall

Most ICF, CMU and wood homes have wood-truss roofs. FEMA’s Hurricane Michael and Ian assessment teams found roof coverings the most commonly damaged envelope element, Irma’s team documented roof-to-wall connection failures, and FEMA’s Marshall Fire team and IBHS identify vents, eaves and the roof as ember entry paths. So two rows score the roof system, and ICF and CMU appear twice:

With a wood-truss roof the roof rows sit at the wood-roof level, whatever the walls are. The concrete walls give the roof straps better anchorage, worth half a point, and nothing more.
With a concrete roof deck tied into the walls the house is a monolithic shell: one load path from roof to footing, no truss uplift joint, a noncombustible unvented deck, and a deck that stops debris. The price is seismic mass, so these systems lose half a point on earthquake, and money, so they lose price and permitting points.
SCIP is scored as a wall-and-roof shell. Its roof is a continuous steel-reinforced shotcrete envelope (typically a 2-inch exterior and 1.5-inch interior wythe, monolithic with the walls), so its roof rows score like the other concrete decks (9), on an engineering basis. No report rates the SCIP roof slab for uplift or roof impact (the approvals’ wind and missile tests cover wall panels), so those rows carry the “Engineered / limited formal testing” badge instead of a lower number.

The exterior face of the wall

One row scores the skin of the walls, apart from their fire rating and structural debris resistance: fire, embers and radiant heat on the face, debris and impact on the finish, and the repair afterwards (see Section 3.8). Grouted CMU with stucco on the block scores 8.5 and SCIP’s reinforced shotcrete faces 8. An ICF wall has about 2.5 inches of EPS outside its concrete, so its score depends on the finish. ICF is scored with the cement-based finish the region picks specify (three-coat stucco on lath, fiber cement or brick veneer): 6.5. Under the thin polymer stucco that is common on ICF, the finish and the foam behind it were holed through by a test missile, and thin finishes over EPS melt back in fire: 4. Wood frame scores 4.5 (combustible sheathing behind the cladding) and SIPs 3. The table below shows where that finish changes the leader.

ICF finish variant: structure-only leader with each ICF finish
Region	ICF + concrete roof, cement-based finish	ICF + concrete roof, thin polymer stucco	Leader, cement-based ICF finish	Leader, thin ICF finish
Florida	8.3	8.0	ICF + concrete roof	CMU + concrete roof
Gulf Coast	8.4	8.2	ICF + concrete roof	CMU + concrete roof
Texas	8.6	8.4	ICF + concrete roof	same
South Atlantic	8.2	8.0	ICF + concrete roof	CMU + concrete roof
Northeast and Mid-Atlantic	8.5	8.2	ICF + concrete roof	same
Northern New England	8.5	8.3	ICF + concrete roof	same
Appalachia	8.5	8.4	ICF + concrete roof	same
Mid-South	8.0	7.9	ICF + concrete roof	same
Lower Midwest	8.6	8.5	ICF + concrete roof	same
Upper Midwest	8.6	8.5	ICF + concrete roof	same
Northern Plains	8.6	8.4	ICF + concrete roof	same
Southern Plains	8.6	8.4	ICF + concrete roof	CMU + concrete roof
Rocky Mountains	8.5	8.2	ICF + concrete roof	CMU + concrete roof
Intermountain West	7.1	6.9	CMU + concrete roof	same
Desert Southwest	8.1	7.8	ICF + concrete roof	CMU + concrete roof
California	7.0	6.8	SCIP	same
Pacific Northwest	6.8	6.7	ICF + concrete roof	same
Alaska	6.2	6.1	Wood frame	same
Hawaii	7.6	7.3	CMU + concrete roof	same
Puerto Rico and US Virgin Islands	7.6	7.3	SCIP	same

Termites and latent moisture

The termite row scores the whole house on two parts, averaged: is the structure food, roof framing included, and can termites reach wood unseen? EPS is not food, but termites tunnel through it from the soil without being seen, so ICF scores 6.5 with a concrete roof and 4.5 under wood trusses, the wood at the end of the foam. SCIP’s foam is encased in concrete on both faces (8, untested); grouted CMU with a concrete roof scores 8.5 and wood frame 3.5. Its weight follows the region’s zone on the IRC termite infestation probability map: very heavy 12%, moderate to heavy 7%, slight to moderate 3%, none to slight 1%. The moisture row averages latent moisture (trapped construction water, drying, condensation, mold), decay and corrosion. The evidence, the zone for every region and our search of insurer positions are in Section 4.8.

Capability by hazard (0–10; the badge is the evidence depth behind the number)
Hazard	ICF + concrete roof	ICF + wood roof	CMU + concrete roof	CMU + wood roof	Wood frame	SCIP	SIPs
Hurricane wind pressure (walls)	9 Extensive data	9 Extensive data	9 Extensive data	9 Extensive data	6 Extensive data	9 Moderate data	6 Moderate data
Windborne debris (9-lb missile)	9 Extensive data	9 Extensive data	9 Extensive data	9 Extensive data	3 Extensive data	9 Moderate data	6 Moderate data
Tornado missile (15-lb, 100 mph)	9 Extensive data	9 Extensive data	9 Extensive data	9 Extensive data	1 Moderate data	4 Emerging data	1 Emerging data
Roof wind performance (uplift, roof-to-wall load path)	9 Moderate data	6.5 Extensive data	9 Moderate data	6.5 Extensive data	6 Extensive data	9 Engineered / limited formal testing	5 Emerging data
Roof fire and ember path	9 Moderate data	5 Extensive data	9 Moderate data	5 Extensive data	5 Extensive data	9 Engineered / limited formal testing	3 Emerging data
Earthquake (SDC D-E, whole-house mass)	5.5 Extensive data	6 Extensive data	5.5 Extensive data	6 Extensive data	9 Extensive data	6.5 Moderate data	3 Extensive data
Fire and wildfire (wall)	9 Moderate data	9 Moderate data	9 Extensive data	9 Extensive data	6 Extensive data	7 Moderate data	3 Moderate data
Flood damage resistance	9 Extensive data	9 Extensive data	6 Extensive data	6 Extensive data	3 Extensive data	6 Moderate data	1 Moderate data
Termites (hidden paths to wood, whole house)	6.5 Moderate data	4.5 Moderate data	8.5 Moderate data	6.5 Moderate data	3.5 Extensive data	8 Emerging data	2.5 Emerging data
Moisture (trapped and latent), decay and corrosion	8 Moderate data	8 Moderate data	7.5 Moderate data	7.5 Moderate data	4 Extensive data	6.5 Emerging data	3.5 Moderate data
Exterior face: fire, embers, debris on the finish, repair	6.5 Emerging data	6.5 Emerging data	8.5 Extensive data	8.5 Extensive data	4.5 Extensive data	8 Moderate data	3 Emerging data

ICF’s wildfire cell notes that its 2–3 hour ratings come from an evaluation report and its EPS faces must be clad. The termite and moisture rows are scored from part-scores in Section 4.8: EPS hides termite galleries; ICF’s concrete dries for years between its foam layers but holds nothing organic; fresh block stores water that harms furring; wood decays above about 20% moisture; SCIP’s thin cover over galvanized mesh is outside every report’s salt-spray scope; and SIP joints have a documented moisture failure record.

The exterior-face row is scored in Section 3.8: the mean of three part-scores (fire on the face, debris on the finish, repair), rounded to the half point. Thin evidence for a finish, such as no wildfire test of ICF with three-coat stucco, shows in the badge only.

Each region’s hazard weights come from the top threats in the region table and the NRI loss shares quoted there. Earthquake and tornado are weighted up for life safety. Inside the wind family the roof gets 40%, debris 25% and wall wind 35%, because the roof is where FEMA finds houses fail first. Where wind is not a top threat the old roof weight moves to the ember row, and 40% of the old fire weight moves there too. Hail, snow, heat, expansive clay, landslide and surge are left out, because the evidence says the roof covering, foundation, siting or elevation decides them, not the structural system. These splits are editorial judgements, and the table is published so they can be checked. The exterior-face row then gets 3% in every region plus 15% of the region’s hurricane-wind and wildfire weight, rounded to the half percent. The termite row gets its zone weight. The other weights are scaled down to make room for both. The exterior face is heaviest in the Rocky Mountains, Florida, the Desert Southwest and Hawaii; termites in the “very heavy” zones.

Hazard weights by region (each row sums to 100%; shown rounded)
Region	Termite zone	Hurricane wind pressure (walls)	Windborne debris (9-lb missile)	Tornado missile (15-lb, 100 mph)	Roof wind performance (uplift, roof-to-wall load path)	Roof fire and ember path	Earthquake (SDC D-E, whole-house mass)	Fire and wildfire (wall)	Flood damage resistance	Termites (hidden paths to wood, whole house)	Moisture (trapped and latent), decay and corrosion	Exterior face: fire, embers, debris on the finish, repair
Florida	Very heavy	16.9%	12.4%	4.4%	19.5%	–	–	–	17.8%	12%	4.4%	12.5%
Gulf Coast	Very heavy	12.8%	9.2%	18.4%	14.7%	–	–	–	18.4%	12%	4.6%	10%
Texas	Moderate to heavy	8%	5.3%	26.7%	8.9%	3.6%	–	5.3%	22.2%	7%	4.4%	8.5%
South Atlantic	Very heavy	13.8%	9.5%	8.6%	15.5%	–	4.3%	–	21.5%	12%	4.3%	10.5%
Northeast and Mid-Atlantic	Moderate to heavy	12.3%	8.8%	4.4%	14.1%	–	–	–	30.8%	7%	13.2%	9.5%
Northern New England	Slight to moderate	8.1%	5.4%	–	9%	1.8%	–	2.7%	40.3%	3%	22.4%	7.5%
Appalachia	Moderate to heavy	1.8%	0.9%	4.4%	1.8%	1.8%	–	2.7%	53.1%	7%	22.1%	4.5%
Mid-South	Moderate to heavy	4.6%	3.7%	32.2%	5.5%	–	18.4%	–	18.4%	7%	4.6%	5.5%
Lower Midwest	Moderate to heavy	4.4%	3.5%	35%	5.3%	–	–	–	30.6%	7%	8.8%	5.5%
Upper Midwest	Slight to moderate	6.4%	4.6%	27.3%	7.3%	–	–	–	31.8%	3%	13.6%	6%
Northern Plains	None to slight	8.2%	5.5%	27.5%	9.2%	1.8%	–	2.7%	18.3%	1%	18.3%	7.5%
Southern Plains	Moderate to heavy	7.7%	5.1%	38.3%	8.5%	3.4%	–	5.1%	12.8%	7%	4.3%	8%
Rocky Mountains	Slight to moderate	8.4%	6.7%	4.2%	10%	13.4%	–	20%	12.5%	3%	8.4%	13.5%
Intermountain West	Slight to moderate	–	–	–	–	15.1%	44.5%	16%	4.4%	3%	8.9%	8%
Desert Southwest	Moderate to heavy	–	–	–	–	24.8%	9.5%	22.9%	14.3%	7%	9.5%	12%
California	Very heavy	–	–	–	–	12.7%	42.4%	12.7%	12.7%	12%	–	7.5%
Pacific Northwest	Slight to moderate	–	–	–	–	6.5%	50.9%	2.8%	13.9%	3%	18.5%	4.5%
Alaska	None to slight	–	–	–	–	1.9%	71.2%	2.8%	4.8%	1%	14.3%	4%
Hawaii	Very heavy	12.5%	8.9%	–	14.3%	3.6%	17.9%	5.4%	–	12%	13.4%	12%
Puerto Rico and US Virgin Islands	Very heavy	13.8%	9.5%	–	15.5%	–	21.5%	–	8.6%	12%	8.6%	10.5%

Limitations

The hazard weights and the dot-to-number mapping are editorial judgements from the cited evidence. They are all published above so they can be checked. Different reasonable weights change close calls. SCIP leads California and Puerto Rico / USVI on the total, both by under 0.1. Hawaii and the Intermountain West go to grouted CMU + concrete roof, and the Pacific Northwest to ICF + concrete roof, because those totals are higher. An earlier edition treated a gap under 0.2 as a tie and let the earthquake row overturn them; that window is retired. The ICF finish assumption still decides Florida, the Southern Plains, the Rocky Mountains and the Desert Southwest when the thin-stucco variant is used.
Evidence depth is a badge, not a score. Where a system has no test for a hazard, its number is engineering judgement from the closest tested assembly (for example SCIP on the 100-mph tornado missile, 4, from a 66-mph pass for a single panel and a 100-mph pass for a double panel), and the badge says “emerging”.
The SCIP roof rows (9 for roof wind and for roof fire and embers) rest on engineering, not on a roof test: a reinforced-concrete section designed and sealed by a registered engineer. That credit holds only for a roof that is actually engineered that way, with the concrete thickness, reinforcement and cover the design calls for, and with an approved roof covering. A roof uplift or roof-impact test would move the badge, not the number.
The ICF exterior-face score assumes a cement-based finish: three-coat stucco on lath, fiber cement or brick veneer. Specify one. A thin polymer stucco or EIFS-type coat directly on the EPS scores 4 instead of 6.5 on that row (see the finish-variant table). No ICF finish, and no SCIP face, has a wildfire wall test (ASTM E2707 / SFM 12-7A-1).
Termite zones are read from the IRC map, whose lines are “approximate only”, and checked against HUD lists and local code tables where we found them; the zone weights are editorial. The ICF termite score assumes termite-resistant EPS or a listed barrier at grade where the code requires it. No SCIP termite test or moisture study exists, so those rows are reasoning from the encased foam and thin faces.
Scores are for the system as a whole house, with the roof it carries, not a specific product. Approvals are product-specific. Check the product’s own evaluation report for your site’s wind speed and Seismic Design Category.
Scores are regional. Overlay zones (flood, surge, tornado, seismic, wildfire, termites) and a site check in the ASCE Hazard Tool and the FEMA flood map can override them.
Personal testimony, including the founder's Ecuador observation, is reported but not scored. Price scores are not quotes; collect real bids for your ZIP code.

How to read this. Section 1 is the region table and the machine-readable state-to-region mapping for the clickable map. Section 2 explains how the picks were made. Section 3 tests the common case for SCIP for hurricane, earthquake and fire zones; it separates code-approval status from demonstrated performance, and keeps the founder's own field observation in a labelled box apart from the cited evidence. Section 4 covers the evidence hazard by hazard. Section 5 gives a short verdict on each building system. Section 6 covers uncertainties and where experts disagree. Section 7 lists every source with its URL.
Bracketed IDs such as [N-55] or [SCIP-08] point to entries in Section 7.

Summary

Most US building losses do not start at the wall.
- FEMA's National Risk Index v1.20 (December 2025 data) ranks inland flooding as the largest expected annual building loss in 45 of 51 states plus DC [N-34, N-36].
- The exceptions are hurricane wind (FL, LA, SC) and earthquake (WA, UT, AK).
- In hurricanes, houses keep failing at the same points: roof coverings, roof decks, openings and water intrusion [FM-07, FM-08, FM-11, FTD-20].
- Storm surge destroys almost anything that is not elevated [FM-06]. Earthquake damage follows mass, cripple walls and soft stories [N-06, N-07, N-09]. Wildfire losses follow embers, vents, eaves, windows and the spacing between houses [N-18, N-22].
- The wall system matters, but in most regions elevation, the roof and the openings matter more. FEMA's assessment teams found roof coverings the most commonly damaged envelope element in Michael and Ian, and documented roof-to-wall connection failures in Irma [FM-07, FM-08, FM-11]. Most ICF, CMU and wood homes still have wood-truss roofs, so this edition scores the whole house: walls, roof and the connection between them (Section 2.4).
The picks rank structure only, for the whole house. "Best" means the strongest demonstrated performance of the walls, the roof and the roof-to-wall connection against the region's hazards: wind and debris, roof uplift, tornado, earthquake, fire and embers, flood, termites (weighted by the IRC termite zone and scored for hidden paths through foam to wood, new on 30 Oct 2026), moisture (now including trapped and latent moisture), decay and corrosion, and the exterior face of the walls. Scores come from test results, ratings and load-path mechanics; how much data stands behind a score is shown separately, as an Extensive, Moderate or Emerging badge beside it. Price and ease of permitting are scored separately; the sliders above the map add them in. Section 2.4 explains the scoring.
- A concrete roof deck tied into concrete walls leads almost everywhere. ICF walls + concrete roof is best in 15 of 20 regions, grouted CMU + concrete roof in Hawaii and the Intermountain West, and SCIP walls + roof in California and Puerto Rico / USVI; wood frame leads Alaska. The same ICF walls under a wood-truss roof score up to 1.1 points lower (0.7 in Florida, 0.5 in Texas, 0.8 in the Rocky Mountains, 1.1 in the Desert Southwest). In the Pacific Northwest the wood-truss roof is the runner-up, 0.07 behind the concrete deck. Only in earthquake-led Alaska does a wood roof lead.
- New on 30 Oct 2026: termites are weighted by zone and scored for hidden paths (Section 4.8). EPS is not termite food, but termites tunnel through it unseen from the soil to wood; Clemson found infestations behind below-grade foam detected only after about five years, and 43% of surveyed pest firms declined to treat such houses [N-76]. The termite row now weighs 12% in the IRC's "very heavy" zones (Florida, the Gulf, Georgia and South Carolina, southern California, Hawaii, Puerto Rico), 7% where "moderate to heavy", 3% "slight to moderate" and 1% "none to slight". Scores: grouted CMU + concrete roof 8.5, SCIP 8 (foam encased in concrete on both faces), ICF + concrete roof 6.5, CMU + wood roof 6.5, ICF + wood roof 4.5 (the foam leads to the trusses), wood frame 3.5, SIPs 2.5. Florida's code bars foam below grade statewide unless protected (FBC Building 2603.8); the 1503.7 and 1816 sections sometimes cited for this are downspout and soil-treatment rules [TM-01, TM-02, TM-03].
- Insurance: we found no US insurer, rating manual or regulator that declines or surcharges ICF homes, over termites or anything else, after repeated searches. Texas law lets insurers give an ICF discount (Insurance Code §2006.002), and one UK lender will not lend on ICF [IN-01, IN-02]. Termite damage is generally a pest-control-contract matter, not insured, and foam that hides termites can void those contracts [TM-12, TM-08]. Section 4.8.
- Latent moisture is now scored inside the moisture row for every system (trapped construction water, drying paths, condensation, mold): ICF's concrete takes years to dry between its foam layers but holds nothing organic and showed no damage in a year-long test; fresh CMU stores water that harms furring; OSB behind stucco concentrates moisture [LM-01, LM-03, LM-04, LM-05]. Row scores: ICF 8, CMU 7.5, SCIP 6.5, wood frame 4, SIPs 3.5.
- New in this edition: the exterior face of the wall is scored for every system (Section 3.8). The row covers fire, embers and radiant heat on the face, debris and impact on the finish, and repair after an event. It weighs 4–13.5% by region, most where hurricanes and wildfire are. Grouted CMU with stucco on the block scores 8.5, SCIP's reinforced shotcrete faces 8, ICF with a cement-based finish (three-coat stucco on lath, fiber cement or brick) 6.5, wood frame 4.5 and SIPs 3. ICF's finish is now an explicit assumption: with the thin polymer stucco that is common on ICF, its EPS face scores 4, and grouted CMU + concrete roof would lead Florida, the Rocky Mountains and the Desert Southwest instead.
- SCIP leads two regions, both close calls on the total. California is 7.11 against 7.04 for CMU + concrete roof, and Puerto Rico / USVI is 7.76 against 7.75. It does not lead the Intermountain West (7.10 against 7.12 for CMU + concrete roof), the Pacific Northwest (6.72 against 6.85 for ICF + concrete roof) or Hawaii (7.87 against 8.05 for CMU + concrete roof). An earlier pass treated a gap under 0.2 as a tie and let the earthquake row (6.5 against 5.5) overturn the higher total. That was a bug: the higher total wins, and the heaviest hazard breaks only an exact tie. The three margins are still small, so a weight change can reverse them. SCIP's roof rows are scored on an engineering basis, and it has no termite, moisture or wildfire test; those gaps show in its badge, not its score.
- Wood frame leads Alaska only (7.6 against 6.6; earthquake is 71% of the weight). No wood-roofed concrete house leads anywhere now: under the termite row a wood-truss roof is the wood that termites reach through ICF's foam.
- SIPs do not lead any region.
- Weighting price and permitting in moves most regions to a wood-truss roof: with structure 50% and price and permitting 25% each, ICF + wood roof leads 11 regions, wood frame 4 (California, Alaska, the Pacific Northwest, the Intermountain West), grouted CMU + wood roof 4 and grouted CMU + concrete roof Florida. With price or permitting alone, FORTIFIED wood frame (plus a safe room where tornadoes are common) leads 16 or 17 regions.
The case for SCIP, tested honestly:
- Termites: agree. SCIP's foam is encased by shotcrete on both faces, so it offers no hidden path from the soil the way exposed EPS does; it scores 8 against 6.5 for ICF with a concrete roof and 8.5 for grouted CMU. No SCIP termite test exists, so the badge is Emerging [N-76, SCIP-04].
- Exterior face: agree. SCIP's 1.25–2 in. of reinforced shotcrete over the foam on both faces is the code's own picture of protected foam (IBC 2603.4.1.1 and the 2603.5.5 exception for foam covered on each face by 1 in. of concrete) [EV-03]. An ICF wall's EPS face depends on its finish: the concrete core stops the missile and the fire, but a thin finish and the foam behind it were holed by a 15-lb missile at about 100 mph, and thin finishes over EPS melt back in fire [EV-08, EV-12]. Grouted CMU with stucco on the block scores higher still (8.5 against 8).
- Hurricane: agree, with conditions. SCIP is the only system sold as a concrete wall-and-roof panel line. Gulf Concrete Technology's FL30050 is the only Florida Product Approval for a SCIP that is valid in the High-Velocity Hurricane Zone (HVHZ), at ±124 psf. Corrected in this edition: the approval lists slab panels, but its wind and missile tests cover wall panels only (PSM 120–240 walls; the 9-lb 2x4 at 80 ft/s on PSM 120), and no SCIP report gives a roof uplift rating or a roof-impact test [SCIP-06, SCIP-08]. (One 3-D wall panel also holds a Miami-Dade NOA, based on a 1994 wall test, with a wood-truss roof in its drawings [SCIP-28].) SCIP's roof is a continuous steel-reinforced shotcrete envelope, so its uplift, debris and roof-fire performance is credited on an engineering basis (an engineer-sealed rational analysis, which the Florida Building Code accepts for uplift design) and scores like the other concrete decks; the missing roof tests show in its evidence badge, not its score [N-106]. On the full Florida hazard mix, ICF with a concrete roof still scores higher, on flood, the tornado missile and durability. SCIP scores 7.9 there, third, behind ICF and grouted CMU with concrete roofs (8.3 each) and ahead of ICF under trusses (7.6).
- Tornado (which overlaps the Gulf and Southeast hurricane belt): not yet. No single-layer SCIP has a published ICC 500 / FEMA P-361 tornado-missile result. ICF does [N-55, ICF-07].
- Earthquake: partly agree (revised in this edition).
  - Code path: ICC-ES ESR-2435 (RSG 3-D/EVG panel) recognises SCIP-type walls in Seismic Design Categories (SDC) C–F when they are engineered as ACI 318 special structural walls [SCIP-27]. GCT's own DrJ report claims SDC B–F; that is a DrJ listing, not an ICC-ES one [SCIP-08]. Two other ICC-ES SCIP reports still stop at SDC B [SCIP-04, N-14].
  - Tests: a full-scale building took 1.2 g on a shake table with only hairline cracks [N-11].
  - Field record: no engineering report records SCIP houses in any real earthquake, including Ecuador 2016 [SCIP-36].
  - Result: the code path and test record put SCIP within reach of the other concrete systems in earthquake regions. It scores 6.5 for earthquake, between the concrete houses (5.5–6) and wood frame (9), with an Emerging-to-Moderate data badge because there is no field record. It leads on structure in California and Puerto Rico / USVI, where it has the highest total, by under 0.1. In the Intermountain West, the Pacific Northwest and Hawaii the higher total belongs to another system; earthquake does not overturn it.
- Fire: partly agree. ASTM E119-tested SCIP walls rate 1 hour (ESR-5623, ESR-3429); ESR-2435 lists 1.5 and 2 hours without citing a test; ratings of 3–4 hours are calculated, not tested [SCIP-04, SCIP-18, SCIP-27, SCIP-12]. A concrete roof removes several ignition paths, but embers, vents, eaves, windows and spacing decide which houses survive, and no SCIP wildfire test exists [N-18, N-22, N-103].
Tally across the 20 map regions (structure only, the map's default):

System	Best in	Runner-up in
ICF walls + concrete roof	15	—
ICF walls + wood-truss roof	—	8
Reinforced CMU walls + concrete roof	2	9
Reinforced CMU walls + wood-truss roof	—	—
FORTIFIED, ignition-resistant or engineered wood frame (wood roof)	1	—
SCIP walls + roof	2	3
SIPs	—	—

With the thin polymer stucco variant for ICF (Section 3.8), grouted CMU + concrete roof leads Florida, the Gulf, the South Atlantic, the Southern Plains, the Rockies, the Intermountain West, the Desert Southwest and Hawaii (8 regions). ICF + concrete roof leads 9, SCIP leads California and Puerto Rico / USVI, and wood frame leads Alaska.

1. Region table

1.1 How the regions map

Every state, DC, Puerto Rico and the US Virgin Islands belongs to exactly one region, so the map can colour whole states (Section 1.4 has the machine-readable list).
Hazards that cross state lines are overlay zones (Section 1.3). An overlay sits on top of a state's colour and overrides the pick inside the zone.
Texas, California, Florida, New York and Washington each contain very different risk regimes. The map shows the regional pick; for a specific site, county or ZIP-level ASCE 7 / USGS / FEMA values can override it. The ASCE Hazard Tool returns site-specific wind, tornado, seismic, snow, ice, rain, flood and tsunami parameters, so it is the right engine for a later address-level version [N-03, M-24].
"Best" means the whole-house system (walls, roof and their connection) with the strongest demonstrated structural and hazard performance against that region's hazard mix: the highest structure score (Section 2.4). Cost, labour, the length of the code path and the amount of evidence do not enter it. They are scored on two separate axes, "best potential price" and "ease of permitting", which the map's sliders can add in. "Runner-up" is the second structure score. Ties within 0.2 points go to the system with the higher score on the region's heaviest-weighted hazard.

1.2 Regions

Region id	States	Top threats	Best system	Runner-up	Reasoning
FL	FL	Hurricane wind (54% of FL building expected annual loss; FL alone is 30% of US hurricane loss), inland flood, storm surge, Formosan and drywood termites, heat and humidity	ICF walls + cast reinforced-concrete roof deck (a monolithic shell), engineered, elevated where mapped, with a cement-based exterior finish (three-coat stucco on lath, fiber cement or brick veneer) and termite-resistant EPS or a listed termite barrier at grade with a 6-in. inspection strip	Fully grouted reinforced CMU walls + concrete roof deck (the regional default; 0.06 behind), ahead of SCIP walls + roof slab from an HVHZ-approved product (today GCT, FL30050) and of ICF with a wood-truss roof	Florida's losses are wind-led, with flood and Formosan termites behind [N-36]. FEMA's assessment teams find the roof the weak point: roof coverings were the most commonly damaged envelope element in Michael and Ian, and Irma showed roof-to-wall connection failures [FM-07, FM-08, FM-11]. So the roof is scored as part of the structure (about 19% of the Florida weight). ICF + concrete roof scores 8.3: concrete walls and a deck tied together resist pressure and the 9-lb debris missile, and ICF adds flood-damage-resistant materials [N-55, ICF-07, FM-19]. CMU + concrete roof scores 8.3, 0.06 behind (grouted CMU has the best documented hurricane field record, Andrew and Irma, but unfilled cells hold floodwater) [FM-01, MAT-02]. Termites carry 12% of the weight in the IRC's "very heavy" zones; Florida's FBC 2603.8 bars foam plastic below grade statewide unless the foam and structure are protected [TM-01, TM-07]. Grouted CMU scores 8.5 on the termite row; ICF + concrete roof scores 6.5, because termites tunnel unseen through EPS, which Clemson's survey linked to infestations found only years later [N-76]. The exterior face carries 12.5% of the weight here: debris and wind strip wall coverings even where the structure holds, and stucco had the lowest damage frequency in Michael [FM-08, EV-28]. Grouted CMU with stucco on the block scores 8.5 on that row and ICF with a cement-based finish 6.5; with the thin polymer stucco common on ICF (4) the ICF house drops to 8.0, behind CMU + concrete roof, so the finish decides Florida (Section 3.8) [EV-08, EV-32]. The same ICF walls under a wood-truss roof score 7.6: the trusses are both the hurricane weak point and the wood that termites reach through the foam (termite row 4.5). SCIP scores 7.9, third, ahead of both wood-roofed concrete houses; its foam is encased in shotcrete on both faces (termite row 8), and the FL30050 wall panels were tested to ±124 psf and the 9-lb missile at 80 ft/s, and its reinforced shotcrete roof is scored like the other concrete decks on an engineering basis; the report publishes no roof uplift or roof-impact test, which shows in the evidence badge, not the score [SCIP-06, SCIP-08, FTD-16]. (RSG's 3-D wall panel holds a Miami-Dade NOA on a 1994 wall test, with a wood-truss roof in its drawings [SCIP-28].) Block is the Florida trade default on price and permitting.
GULF	LA, MS, AL	Hurricane wind (LA 57% of loss; #2 in MS and AL), surge (Katrina, Ida), inland flood, Dixie Alley tornadoes (rising), Formosan termites, humidity; weak code adoption in MS and AL	ICF walls + concrete roof deck (flat wall, ≥6-in. core, cement-based exterior finish, termite-resistant EPS or a listed barrier at grade)	Fully grouted reinforced CMU walls + concrete roof deck, ahead of ICF walls + wood-truss roof built with FORTIFIED Gold roof details	Hurricanes and violent tornadoes overlap here. ICF has hurricane- and tornado-missile test results: FEMA P-361 accepts uniform ICF cores of 4 in. or more for safe rooms [N-55, ICF-07]. Fully grouted reinforced CMU also stops the 100-mph missile, but unfilled cells hold floodwater [CMU-01, FM-19]. A concrete deck removes the roof as a failure point (ICF + concrete roof 8.4). CMU + concrete roof (8.3) is the runner-up, ahead of ICF under a FORTIFIED-grade wood-truss roof (7.8): stucco on grouted block scores 8.5 against 6.5 for a cement-based finish over ICF's foam on the exterior-face row (10% here), and the Gulf is a "very heavy" Formosan termite zone (12% of the weight), where EPS hides galleries to the roof framing [EV-28, EV-08, N-76, N-77]. SCIP scores 7.2: strong hurricane wall and a concrete roof slab, but its single-layer wall passed the 15-lb missile only at 66 mph [SCIP-26]. FORTIFIED Gold wood frame plus a safe room has the best real-storm loss data in the country (Hurricane Sally: loss frequency down 55–74%) and leads on price and permitting [FTD-09, COST-16, N-44, M-12], but the wall itself has no debris resistance [CMU-04].
TX	TX	Hail (#1 US, 30% of US hail loss), tornado (#1 US), inland flood (#2 US), hurricane wind (16% of TX loss), expansive clays, wildfire; no mandatory statewide residential code	ICF walls + concrete roof deck on a post-tensioned slab, cement-based exterior finish, termite-resistant EPS or a listed barrier at grade	Fully grouted reinforced CMU walls + concrete roof deck, ahead of ICF walls + wood-truss roof with an impact-rated roof covering (FORTIFIED Roof + Hail Supplement) [N-62]	Texas losses are led by tornado, flood and hail, and roof coverings take most hail impacts [N-36, N-61]. Tornado is the life-safety hazard, and ICF is the wall with independent tornado-missile results [N-55, ICF-07]; it also uses flood-damage-resistant materials [FM-19]. Fully grouted CMU stops the same missile; partly grouted CMU does not [CMU-01, CMU-04]. A concrete deck lifts the roof rows (ICF + concrete roof 8.6 against 8.1 under trusses), but a concrete deck does not stop hail damage to its covering. Grouted CMU + concrete roof (8.2) is now second: most of Texas is a "moderate to heavy" termite zone, very heavy on the Gulf coast (7% of the weight), and termites can reach a wood-truss roof unseen through ICF's foam [N-76, TM-07]. Hail and debris also mark wall finishes; the exterior-face row carries 8.5% of the weight here. Post-tensioned slabs are the proven answer to the clays, whatever the wall [N-72]. About 95% of tornadoes are EF0–EF2, which code-plus wood framing survives; Moore's 135-mph upgrades cost about $1,500–2,500 per home, and a safe room covers the violent tail [N-55, N-56, N-58]. That FORTIFIED wood + safe-room package leads on price and permitting.
SATL	GA, SC, NC	Hurricane wind (SC 48%, #1; NC 32%; GA 18%), inland flood (Helene: 77 direct freshwater-flood deaths in NC), surge, tornado, Charleston earthquake zone, Formosan termites; NC code lag	ICF walls + concrete roof deck	Fully grouted reinforced CMU walls + concrete roof deck, ahead of ICF walls + wood-truss roof (FORTIFIED Gold roof)	Coastal SC and NC take major hurricanes, flood is the largest loss after wind, and the Charleston area also needs seismic design. ICF has strong hurricane, tornado-missile and flood results and can be engineered for any SDC under ACI 318 [N-55, M-21, FM-19]. ICF + concrete roof scores 8.2. The concrete deck is worth about 0.6 points over trusses (ICF + wood roof 7.6), because the roof is a leading hurricane failure point and, in Georgia and South Carolina's "very heavy" termite zone (12% of the weight), the wood that termites reach through the foam [FM-08, FM-11, TM-07, N-76]. Grouted CMU + concrete roof (8.0) is second: it matches ICF on wind and debris and beats it on the exterior face (8.5 against 6.5) and termites (8.5 against 6.5), but not on flood. SCIP is fourth (7.5). One SCIP-type panel now has an ICC-ES path in the Charleston zone (ESR-2435) [SCIP-27], but its single-layer wall passed the missile only at 66 mph. North Carolina has paused code adoption, so the structure has to make up for code gaps [N-41, M-11]. FORTIFIED Gold wood frame leads on price and permitting, and is a defensible choice inland, where design wind speeds are moderate.
NEMA	NJ, NY, CT, RI, MA, DE, MD, DC, VA	Inland and coastal flood (NJ #2 US coastal flood), strong wind (NJ #1 US), nor'easters and hurricane surge (1938, Sandy), snow (ASCE 7-22 raised many mid-latitude loads, e.g. Baltimore 25→60 psf), freeze-thaw	ICF walls + concrete roof deck (basement and above grade)	ICF walls + wood-truss roof	The hazard mix is flood, coastal storm and freeze-thaw. ICF answers all three: debris-resistant concrete, flood-damage-resistant materials under NFIP TB-2, and continuous EPS insulation. It also suits the region's basement culture [FM-19, ICF-11, N-63]. The concrete deck scores 8.5 against 8.0 for trusses on roof wind performance and the unvented, noncombustible deck, and because in this "moderate to heavy" termite zone (7% of the weight) a wood-truss roof is what termites reach through the foam [N-76]. Grouted CMU is as strong in wind, but unfilled cells hold water and block needs freeze-thaw detailing [FM-19, N-67]. FORTIFIED wood frame leads on price and permitting, and strong code programmes (IBHS ranks VA 1st and NJ 4th) narrow the structural gap inland [M-11].
NNE	VT, NH, ME	Inland flood (VT 2023), snow and ice, deep cold (IECC 6–7), coastal storms (ME)	ICF walls + concrete roof deck (foundation and walls)	ICF walls + wood-truss roof (the usual snow-load roof)	Inland flood is the largest loss, and cold, wet winters test durability. ICF is flood-damage-resistant and does not rot, on the frost-protected or basement foundations that are already standard [FM-19, N-65]. A concrete deck scores 8.5 against 8.2 for trusses, but snow load is decided by roof framing [N-64], so most houses here will keep trusses and the ICF-with-trusses house is the realistic runner-up. Grouted CMU is as durable structurally but needs added insulation. SIPs give the best whole-wall R-values and very low air leakage, and there are regional makers [SIP-15, SIP-16], but energy is not a structural score: OSB skins fail NFIP flood rules, and the Juneau SIP roof failures came from air leakage at unsealed joints [FM-19, N-79]. SIPs and wood frame lead on price and permitting.
APP	PA, WV	Inland flood (PA #6 US), flash flood and landslide in Appalachian valleys, snow, freeze-thaw; WV has no mandatory code	ICF walls + concrete roof deck (basement to roof)	ICF walls + wood-truss roof	Flash flooding in narrow valleys is the dominant loss, so siting and elevation are the first defence [N-40]. Where a house must stay near water, concrete and closed-cell foam are on FEMA's flood-damage-resistant list, while OSB and paper-faced gypsum are not [FM-19]. Wind and fire barely weigh here, so the roof matters little: ICF with a concrete deck (8.5) and with trusses (8.2) are close. Concrete block is on the flood list too, but unfilled cells hold water. Elevated FORTIFIED wood frame with flood-damage-resistant materials leads on price and permitting.
MIDS	MO, AR, TN, KY	Tornado (rising in AR, MO, TN, KY), inland flood, ice storm (AR #1, MO #4), hail, New Madrid earthquake (TN: 24% of loss); low code adoption	ICF walls + concrete roof deck, engineered under ACI 318 in New Madrid SDC D, plus a safe room for violent tornadoes	ICF walls + wood-truss roof (8.0 against 7.8)	This is the region where tornado and earthquake hazards overlap most. Tornado is the governing life-safety hazard, and ICF and fully grouted CMU are the walls with tornado-missile results [N-55, CMU-01]. A concrete deck is heavier, which costs it on the 20% earthquake weight, and it gains on roof wind performance; the two nearly cancel, so the ICF house with a wood-truss roof stays close; termites (7% of the weight, "moderate to heavy") tip it further to the concrete roof [N-76, TM-07]. ICF needs an engineer in SDC D because IRC R608 stops at SDC C [N-04, N-05]. Light frame keeps seismic mass low (code dead loads: 15 psf for a wood wall against 85 psf for 6-in. concrete) and, with a safe room, leads on price and permitting [N-04, N-55]. SCIP has an ICC-ES path in SDC D (ESR-2435) but a 66-mph missile result for the single-layer wall, so it ranks well below the other concrete systems [SCIP-27, SCIP-26].
LMW	IA, IL, IN, OH	Inland flood (IL #3 US), tornado (IL #2, OH #4, IN #5, IA #7 in tornado loss), hail, winter storms; patchy code adoption	ICF walls + concrete roof deck (with an ICF basement or safe-room core)	ICF walls + wood-truss roof	Tornadoes and inland flooding lead the losses. ICF has tornado-missile and flood-damage-resistance results [N-55, ICF-07, FM-19]; grouted CMU matches it on the missile but not on flood. The roof is the first thing a tornado removes, and a concrete deck tied to the walls keeps it (8.6 against 8.3 for trusses), though only a safe room protects against EF3+. Tornadoes here are mostly EF0–EF2, so Moore-style connection upgrades on wood frame plus a shelter give the best risk reduction per dollar and lead on price and permitting [N-55, N-56, N-58].
UMW	MN, WI, MI	Inland flood, tornado (MI #3 in tornado loss), hail (WI, MN top 8), strong wind, deep cold (IECC 6–7), snow	ICF walls + concrete roof deck (basement to roof)	ICF walls + wood-truss roof	Flood, tornado and hail exposure favour a concrete-core wall, on the frost-protected or basement foundations that are already standard [N-65]. ICF adds continuous insulation; grouted CMU needs it added and careful freeze-thaw detailing [N-67]. The concrete deck scores 8.6 against 8.4 for trusses. SIPs are the light, efficient alternative but give no debris protection.
NPL	ND, SD	Winter storms (ND #1 US), hail (15% of ND loss), tornado, extreme cold (IECC 6–7), expansive shale; SD has no mandatory code	ICF walls + concrete roof deck	ICF walls + wood-truss roof	The logic is the same as UMW: tornado and severe wind favour a missile-resistant wall, hail wrecks siding but not concrete-faced walls, and ICF basements are common. A concrete deck scores 8.6 against 8.3 for trusses. Termites barely count here ("none to slight", 1%). Foundations on swelling clays and shales need engineering whatever the wall [N-70].
SPL	OK, KS, NE	Tornado, hail, ice storm (OK #2 US), strong wind, grassland fire, expansive clays (OK); weakest code adoption in the country	ICF walls + concrete roof deck with an ICF safe-room core, a cement-based exterior finish and termite protection at grade	Fully grouted reinforced CMU walls + concrete roof deck, ahead of ICF walls + wood-truss roof with a FORTIFIED Roof + Hail Supplement covering	This is the tornado core of the country. ICF and fully grouted CMU are the walls with tornado-missile results; partly grouted CMU was perforated at 104 mph [N-55, ICF-07, CMU-01, CMU-04]. ICF + concrete roof scores 8.6; CMU + concrete roof 8.5, second because hail and wind-borne debris on the wall finish (the exterior-face row, 8% here) favour stucco on block over a finish on foam, and termites ("moderate to heavy", 7%) favour block with no foam; ICF under trusses 8.1. With the thin polymer stucco variant on ICF, CMU + concrete roof leads. Moore's 135-mph code and the Tuscaloosa and Joplin damage surveys show that designing wood frame for EF0–EF2 and sheltering for EF3+ cuts damage across up to 85% of an affected area [N-56, N-58]. That package, with FORTIFIED's third-party verification and Oklahoma's Roof + Hail grants, leads on price and permitting [N-41, N-45].
ROCK	CO, WY	Hail (CO #2 US), wildfire (CO #3 US; Marshall Fire), snow, wind, Front Range expansive soils, cold; Colorado Wildfire Resiliency Code applies in the wildland-urban interface (WUI)	ICF walls + concrete roof deck (Class A, unvented, with ember-resistant eaves and openings), with a cement-based exterior finish (three-coat stucco on lath, fiber cement or masonry), not a thin polymer coat on the foam	Fully grouted reinforced CMU walls + concrete roof deck (within 0.2; the leader with a thin-coat ICF finish)	Hail and grass-to-house fire are the two big losses. The Marshall Fire spread largely house to house, and FEMA recommends noncombustible cladding and Class A roofs where houses are within 30 ft of each other [N-21]. FEMA's Marshall team and IBHS identify vents, eaves and roofs as the ember entry paths [N-103, M-13]. A concrete roof deck removes the vented attic and the combustible deck, which is why both concrete-roofed systems (ICF 8.5, CMU 8.4) lead the same walls under trusses (ICF 7.7). The exterior face weighs most here (13.5%): FEMA found concrete block and brick "remained standing more often" in the Marshall Fire and flagged exterior rigid foam for study, and thin finishes over EPS melt back and pool-burn in full-scale tests [EV-13, EV-12]. ICF walls carry ICC-ES-listed 2–3 h fire ratings, but from the interior side; their EPS faces need a noncombustible covering, and Colorado's Class 2 walls limit coverings to noncombustible, fire-retardant-treated or ignition-resistant materials [M-21, EV-01, N-29]. So ICF leads only with a cement-based finish (exterior face 6.5 against 8.5 for grouted CMU); with a thin polymer stucco it scores 8.2 and grouted CMU + concrete roof leads (Section 3.8). Grouted CMU is fully noncombustible but less flood- and cold-tolerant. Neither replaces vents, eaves, glazing and Zone 0 [N-22]. A SCIP shell with a concrete roof scores 7.6: no wildfire test, and the cold climate exposes truss-wire thermal bridging [N-68]. Ignition-resistant FORTIFIED wood frame (CWRC Class 2 walls) leads on price and permitting.
IMW	UT, ID, MT, NV	Earthquake (UT: 46% of loss, #1), wildfire (NV, ID, UT, MT high to very high risk to homes), snow, cold, dry heat	Fully grouted reinforced CMU walls + concrete roof deck (highest total, 7.12; SCIP is 0.02 behind and is not promoted on earthquake)	SCIP walls + roof (7.10), ahead of ICF walls + concrete roof deck (7.08)	This is a classic fire-plus-earthquake zone. CMU + concrete roof (7.12) leads SCIP (7.10) and ICF + concrete roof (7.08). The gap is small, and the higher total wins. SCIP scores 6.5 on earthquake against 5.5 for the concrete-walled, concrete-roofed houses, which would decide only an exact tie. It has an ICC-ES path in SDC C–F (ESR-2435), took 1.2 g on a full-scale shake table with hairline cracks only, weighs about 35 psf against 80–85 psf, and its concrete roof slab removes the vented attic [SCIP-27, N-11, SCIP-08, N-18]. The concrete-roofed systems win the fire and ember weight (a noncombustible deck, 9 against 5 for any wood roof) [N-04, M-21]. Ignition-resistant wood frame (6.7) still scores 9 for earthquake and is prescriptive through SDC D2, and Salt Lake City has among the highest US design ground motions [N-02, N-04, N-06, N-08]; it led the previous edition on the tie rule but now sits 0.4 behind the leaders, outside the tie window, because the exterior-face row (8% here) counts its combustible sheathing behind the cladding (4.5 against 8 for SCIP's shotcrete faces) [EV-12, EV-16]. Wildfire survival still depends on vents, eaves, glazing and spacing [N-18, N-22, N-103]. SCIP has no wildfire test and no real-earthquake record.
DSW	AZ, NM	Wildfire (AZ #2 US), extreme heat, monsoon flash flood, expansive soils, desert subterranean termites; mostly low-to-moderate seismic hazard	ICF walls + concrete roof deck, with a cement-based exterior finish and termite protection at grade	Fully grouted reinforced CMU walls + concrete roof deck (the leader with a thin-coat ICF finish)	Wildfire is the heaviest hazard, with flood, termites and some seismic hazard behind. The concrete roof deck matters most here: embers enter through roofs and vents, and a noncombustible unvented deck scores 9 against 5 for a wood roof (ICF + concrete roof 8.1, ICF + trusses 7.0) [N-18, N-103, M-13]. ICF scores above CMU on flood (9 against 6), while grouted CMU is fully noncombustible with no foam on its face (exterior face 8.5 against 6.5 for ICF with a cement-based finish), so CMU + concrete roof is close behind (8.0). With a thin polymer stucco on the EPS, ICF drops to 7.8 and CMU leads (Section 3.8) [M-21, EV-12]. The desert is a "moderate to heavy" termite zone (7%): grouted block has no foam to hide galleries (termite row 8.5 against 6.5 for ICF with a concrete roof) [N-76, TM-13], and ORNL ranks Phoenix among the most favourable climates for massive walls [N-69, N-78]. SCIP scores 7.45, third: its shotcrete faces leave no foam exposed, and an unvented insulated concrete roof suits desert heat and removes attic vents, but that is our inference and has not been tested, and its R-values are calculated [N-18, SCIP-08, N-68]. Seismic design no longer limits it (ESR-2435 covers SDC C–F) [SCIP-27]. AAC suits the climate too, but US supply is uncertain [N-84–N-86].
CA	CA	Inland flood (#1 US), earthquake (58% of US earthquake loss), wildfire (46% of US wildfire loss), coastal flood; insurance stress (FAIR Plan 668,609 policies, Dec 2025)	SCIP walls + roof slab under an evaluation report that covers the site's SDC (ESR-2435), with the 2025 California WUI Code package (Class A roof, ember-resistant vents, enclosed eaves, tempered dual-pane glazing, Zone 0, spacing); the highest total by 0.07	Fully grouted reinforced CMU walls + concrete roof deck, engineered, tied with ICF walls + concrete roof deck under ACI 318 with a cement-based exterior finish and termite protection (both within 0.1; ICF with a wood-truss roof and engineered wood frame within 0.8)	Earthquake is the heaviest hazard, with wildfire, flood and termites behind. Light, ductile wood frame has the best earthquake record when cripple walls and soft stories are avoided [N-04, N-07, N-09]. Engineered concrete walls carry more mass but add fire, flood and termite resistance that wood frame lacks, and a concrete roof deck removes the ember path through the attic. SCIP (7.11), CMU + concrete roof (7.04) and ICF + concrete roof (7.03) are within 0.1 of each other, a close call; SCIP now has the highest total outright. It scores 6.5 for earthquake against 5.5 for the solid concrete houses, weighs about 35 psf against 85 psf, has an ICC-ES path in SDC D–E (ESR-2435) and took 1.2 g on a full-scale shake table with hairline cracks only [SCIP-27, N-11]. Southern California is in the IRC's "very heavy" termite zone (12% of the weight here; moderate to heavy in the north) [TM-13]: SCIP's foam is encased in shotcrete on both faces (termite row 8), grouted CMU has none (8.5), while ICF's EPS hides galleries (6.5 with a concrete roof, 4.5 under trusses) [N-76]. On the exterior-face row (7.5%) SCIP scores 8 against 6.5 for ICF with a cement-based finish and 4 with a thin polymer stucco. California's WUI walls must be noncombustible, ignition-resistant, 1-hour rated from the exterior side or pass ASTM E2707 / SFM 12-7A-1; no ICF, EIFS or SCIP assembly appears in the State Fire Marshal's listed-products handbook, and ICF's published ratings are interior-side unless at least 1 in. of concrete or masonry covers the foam [EV-05, EV-06, EV-01]. SCIP has no real-earthquake record, no wildfire test, no termite test and no published roof uplift test, and ESR-5623 and ESR-3429 panels remain limited to SDC A–B [SCIP-04, N-14]. CMU + wood roof (6.5), ICF + wood roof (6.5) and wood frame (6.35) follow. Wood frame leads on price and permitting: it is code-prescriptive, while concrete must be engineered beyond SDC C. In the 2025 LA fires, 76–89% of surveyed homes already had noncombustible siding; spacing, Zone 0 fuel and the number of hardened features governed survival [N-22].
PNW	WA, OR	Earthquake (WA 50% of loss, #1; OR 42%), inland flood, wildfire east of the Cascades, wet marine climate (moisture, decay), coastal tsunami	ICF walls + concrete roof deck (highest total, 6.85)	ICF walls + wood-truss roof (6.78), ahead of SCIP walls + roof (6.72)	The seismic logic is the same as California's, with Cascadia M9 potential [N-02]. ICF + concrete roof (6.85) leads ICF + wood roof (6.78) and SCIP (6.72). The gaps are small, and the higher total wins. SCIP scores 6.5 on earthquake against 6 under trusses and 5.5 under a concrete roof, which would decide only an exact tie [SCIP-27, N-11]. SCIP gained here because the moisture row now scores latent moisture explicitly (6.5 against 6 before; its thin faces dry to each side, though no SCIP moisture study exists) and because grouted CMU and ICF lost a little on the termite row. Rain-driven moisture is the chronic hazard: wood scores 4 against 8 for ICF, because wood is safe from decay only below about 20% moisture content and OSB behind a reservoir cladding concentrates moisture, while ICF's concrete holds nothing organic but takes years to dry between its foam layers [MAT-07, N-80, LM-05, LM-01]. Wood frame (6.53) scores 9 on earthquake but sits 0.3 behind, outside the tie window. Washington rescinded its state WUI amendments in 2026, so WUI detailing is voluntary in many places [N-30]. Termites barely count west of the Cascades ("slight to moderate", 3%).
AK	AK	Earthquake (83% of AK loss), extreme cold (IECC 7–8), interior wildfire, coastal erosion; no statewide residential code	Engineered light wood frame with a wood roof, super-insulated	SCIP walls + roof under an evaluation report that covers the site's SDC (ESR-2435) (ICF walls + wood-truss roof 0.1 behind)	Earthquake is 83% of building loss [N-36]. After the 2018 M7.1 earthquake, modern Anchorage homes had mostly cosmetic damage; communities with poor code enforcement had 18–20 times more [N-08]. A wood-roofed system wins clearly: 7.6 against 6.6, because earthquake carries 71% of the weight and wood frame scores 9 for it, while wind and wildfire weigh almost nothing here and so a concrete roof earns little. Wood is light, locally buildable and easy to ship. ICF is engineerable for the demand and durable, but heavy. SIPs suit prefabrication but are outside IRC seismic limits and need flawless joint sealing in extreme cold [SIP-02, N-79]. SCIP (6.6) is second, ahead of ICF with a wood roof (6.5): it is lighter, and its shotcrete faces score higher on the small exterior-face weight. It has no local supply and truss-wire bridging matters in IECC 7–8 [N-68].
HI	HI	Hurricane (Iniki 1992), earthquake (Hilo among the highest US design motions), wildfire (Lahaina 2023), Formosan termites ("very heavy"), salt-air corrosion	Fully grouted reinforced CMU walls + concrete roof deck (highest total, 8.05; the island trade default)	SCIP walls + roof (7.87), ahead of ICF walls + concrete roof deck, with no foam below grade in 'very heavy' termite areas	Termites, salt, fire, hurricane and earthquake all count here, and the roof is a hurricane and ember path. CMU + concrete roof scores 8.05 and SCIP 7.87. The higher total wins. SCIP scores 6.5 on earthquake against 5.5 (Hilo has among the highest US design motions), and that would decide only an exact tie [N-02, SCIP-27]. This is a close call that a small change in weights can reverse. Hawaii is "very heavy" for Formosan termites (12% of the weight) [TM-14, N-77]: grouted CMU scores 8.5 on the termite row, SCIP 8 (foam encased in shotcrete on both faces, untested) and ICF 6.5, because EPS hides galleries and foam below grade is restricted [N-74, N-76]. ICF + concrete roof scores 7.6. Stucco direct on grouted block has no foam to melt or puncture, which gives CMU the best exterior-face row (12% here). Grouted CMU is the island trade default, with IRC seismic provisions to SDC D2, and leads on price and permitting [N-04]. SCIP's roughly 1-in. mesh cover is thin for salt spray, and salt-spray durability is outside every SCIP report's scope, which its moisture-and-corrosion row reflects (6.5 against 7.5 for CMU) [N-14, SCIP-11].
PRVI	PR, VI	Hurricane (PR #3 US hurricane loss; Maria), earthquake (2020 sequence), landslide, flood	SCIP walls + roof, engineered for SDC D under an evaluation report that covers it (the highest total by 0.01, and ahead on earthquake)	Reinforced CMU walls + concrete roof slab (tied; the island default), ahead of ICF or cast reinforced-concrete walls + concrete roof slab with a cement-based exterior finish	Hurricane and earthquake are equally weighted here. Concrete-walled houses handle both if detailed: the 2020 collapses were mostly houses on slender columns with open ground floors, not concrete-walled houses [N-98]. A concrete roof slab is the local norm and the roof is the leading hurricane failure point. SCIP (7.76), CMU + concrete roof (7.75) and ICF + concrete roof (7.60) are within 0.2, and SCIP also leads on earthquake, the heaviest hazard (6.5 against 5.5). Puerto Rico is in HUD's "very heavy" termite zone (12% of the weight) [TM-07]: grouted CMU scores 8.5 on the termite row, SCIP 8 and ICF 6.5, which moved ICF from first to third. SCIP's shotcrete faces also score above ICF's finished foam on the exterior-face row (10% here), where FEMA's Maria team saw the stucco coat on a concrete hospital crack, spall and admit water [EV-41]. ICF under trusses scores 7.1. SCIP's roof rows are scored on an engineering basis, like the other concrete decks; no SCIP roof uplift or roof-impact test has been published, which its badge shows. ICF leads CMU on flood-damage resistance [FM-19]. Islanders already build in concrete block with slab roofs, so CMU leads on price and permitting. SCIP's lighter shell is attractive: GCT's report lists the Puerto Rico Building Code, its parent company is in Puerto Rico, and an ICC-ES path to SDC D–E now exists [SCIP-08, SCIP-33, SCIP-27]. It has no documented record from Maria or the 2020 earthquakes.

Notes on the figures:

Loss shares and ranks are building-value Expected Annual Loss from FEMA NRI v1.20 (state layer, "December 2025") [N-36]. Ranks include territories.
In the NRI, "hurricane" means wind only; surge is counted under coastal flooding [N-34].
Code-adoption statements come from FEMA BCAT [N-41] and IBHS Rating the States 2024 [M-11].

1.3 Overlay zones (these override the state colour inside the zone)

Overlay id	Where (data layer)	Override
OV-HVHZ	Miami-Dade (FIPS 12086) and Broward (12011) counties, FL: the Florida Building Code High-Velocity Hurricane Zone	Best: a concrete shell with a concrete roof. Wind, debris and the roof weigh more here than in the rest of Florida, which narrows the gap between the three concrete shells; the map's scoring does not model the overlay separately. The options are ICF or fully grouted reinforced CMU with a cast concrete roof deck, or SCIP walls + roof with an HVHZ product approval. GCT's FL30050 is the only Florida Product Approval for a SCIP valid in the HVHZ, and the only HVHZ approval found that lists slab panels. Its wind and missile tests cover the wall panels only; no uplift rating or roof-impact test for a slab appears in the TER [SCIP-08]. FL46983 (FSS, formerly marketed as Concrewall USA) and Schnell's ESR-3429 are not HVHZ-approved [SCIP-06, SCIP-01, SCIP-04, N-14]. GCT, FSS, Schnell and Emmedue hold no Miami-Dade Notice of Acceptance (NOA) [SCIP-24]. The one 3-D panel NOA found is RSG's NOA 25-0421.03 (wall panel system with a wood-truss roof in its drawings; large and small missile; based on a 1994 wall test) [SCIP-28]. Confirm local acceptance with the building official. FBC 2023 §1626.4 deems exterior reinforced, solid normal-weight concrete (no voids) at least 2 in. thick, and 8-in. CMU, to meet the HVHZ impact rule without a missile test, so impact resistance alone does not separate the concrete systems; whether a foam-core SCIP roof with a 2-in. exterior wythe qualifies is the building official's call (Section 3.1) [N-106, N-96]. Any HVHZ concrete roof still needs a product-approved roof covering (§1514, §1515) [N-106].
OV-VZONE	FEMA Coastal High Hazard Areas (V zones) and Coastal A zones (FEMA National Flood Hazard Layer)	Elevation on open foundations (piles or columns) with breakaway walls below the base flood elevation is mandatory and comes before any wall choice [N-37, M-01]. Above that elevation, apply the regional pick. Lighter shells (wood, SCIP) put less load on piles than ICF or CMU, and a concrete roof adds to it.
OV-HURRCOAST	Atlantic and Gulf windborne-debris regions (ASCE 7 basic wind speed ≥140 mph, or ≥130 mph within 1 mile of the coast)	Upgrade to a concrete shell (ICF, SCIP or grouted CMU) with a concrete roof deck; a FORTIFIED Gold wood-truss roof is the lower-cost step. Inland of this line, FORTIFIED wood frame is the lower-cost option where the regional structural pick is not affordable. Risk Category II design speeds on the Gulf and Florida coast run roughly 140–180 mph (180 mph in the Keys) [M-06, M-07].
OV-TORN250	FEMA P-361 250-mph tornado safe-room design zone (central and southern US)	Add an ICC 500 / FEMA P-361 safe room to any system. FEMA-funded residential safe rooms must be designed for 250-mph tornado wind and missile loads [M-05]. FEMA reports no failure of a safe room built to its criteria [N-55].
OV-NMSZ	New Madrid seismic zone (MO bootheel, NE AR, W TN incl. Memphis, W KY, S IL) and the Charleston, SC zone (USGS NSHM 2023 / FEMA SDC maps)	Engineer any concrete system under ACI 318 (IRC R608 stops at SDC C) [N-04]. Light frame + safe room is the shorter code path and carries less seismic mass. Of the ICC-ES SCIP-type reports, only ESR-2435 reaches SDC C–F, as special structural walls; ESR-5623 and ESR-3429 stop at SDC B [SCIP-27, SCIP-04, N-14].
OV-WUI	CA Very High / High Fire Hazard Severity Zones, Colorado CWRC areas, Oregon R327 areas, other mapped WUI	Apply the WUI package on top of the regional system: Class A roof; ember-resistant vents (1/16–1/8-in. mesh or ASTM E2886); enclosed eaves; tempered multi-pane glazing; noncombustible bottom 6 in. of wall; Zone 0. Keep houses ≥30 ft apart where possible [N-21, N-22, N-25, M-13]. Concrete walls help most where houses are <30 ft apart.
OV-TERMITE	IRC "very heavy" termite probability areas (2018 IRC Figure R301.2(7); 2021 Figure R318.4; 2024 Figure R305.4): Florida, the Gulf Coast, Georgia and South Carolina, southern California, Hawaii and Puerto Rico; Formosan range up the Atlantic to Virginia	No foam plastic on the exterior face or under foundation walls or slabs below grade, and 6 in. of clearance between above-grade foam and soil, unless an exception applies: an all-noncombustible or treated-wood structure including the roof, an approved protection method (for example termite-resistant EPS or a listed barrier), or the inside of a basement wall (IRC R318.4 in 2009–2021; R305.4 in 2024; FBC Building 2603.8 and FBC Residential R318.4 in Florida, where 2603.8 applies statewide) [N-74, N-75, TM-01, TM-06]. FBC 1503.7 (downspouts) and FBC 1816 (soil treatment) are termite rules but not foam rules [TM-02, TM-03]. Above grade the scores already count the hidden-path risk (Section 4.8); this overlay adds the foundation rules.
OV-CLAY	Expansive-clay belts (central TX, Front Range CO, OK, Dakotas, parts of CA) (USGS I-1940)	The foundation (post-tensioned slab or engineered piers) governs [N-72]. Avoid brittle finishes. Reinforced concrete walls tolerate movement about as well as stud frames [N-73].
OV-TSUNAMI	Pacific tsunami inundation zones (WA, OR, N. CA, AK, HI)	Siting comes before the building system. No residential system is designed to resist a tsunami.

1.4 Machine-readable mapping

State → region (50 states + DC + PR + VI; every key appears exactly once):

Wider than the screen. Scroll this block sideways.

{
  "AL": "GULF", "AK": "AK",   "AZ": "DSW",  "AR": "MIDS", "CA": "CA",
  "CO": "ROCK", "CT": "NEMA", "DE": "NEMA", "DC": "NEMA", "FL": "FL",
  "GA": "SATL", "HI": "HI",   "ID": "IMW",  "IL": "LMW",  "IN": "LMW",
  "IA": "LMW",  "KS": "SPL",  "KY": "MIDS", "LA": "GULF", "ME": "NNE",
  "MD": "NEMA", "MA": "NEMA", "MI": "UMW",  "MN": "UMW",  "MS": "GULF",
  "MO": "MIDS", "MT": "IMW",  "NE": "SPL",  "NV": "IMW",  "NH": "NNE",
  "NJ": "NEMA", "NM": "DSW",  "NY": "NEMA", "NC": "SATL", "ND": "NPL",
  "OH": "LMW",  "OK": "SPL",  "OR": "PNW",  "PA": "APP",  "RI": "NEMA",
  "SC": "SATL", "SD": "NPL",  "TN": "MIDS", "TX": "TX",   "UT": "IMW",
  "VT": "NNE",  "VA": "NEMA", "WA": "PNW",  "WV": "APP",  "WI": "UMW",
  "WY": "ROCK", "PR": "PRVI", "VI": "PRVI"
}

Region → best / runner-up (short labels for map tooltips):

Wider than the screen. Scroll this block sideways.

{
  "FL": {"best": "ICF + concrete roof deck", "runner_up": "CMU + concrete roof"},
  "GULF": {"best": "ICF + concrete roof deck", "runner_up": "CMU + concrete roof"},
  "TX": {"best": "ICF + concrete roof, PT slab", "runner_up": "CMU + concrete roof"},
  "SATL": {"best": "ICF + concrete roof", "runner_up": "CMU + concrete roof"},
  "NEMA": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "NNE": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "APP": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "MIDS": {"best": "ICF + concrete roof (SDC D) + safe room", "runner_up": "ICF + wood-truss roof"},
  "LMW": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "UMW": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "NPL": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "SPL": {"best": "ICF + concrete roof + safe-room core", "runner_up": "CMU + concrete roof"},
  "ROCK": {"best": "ICF (cement-based finish) + Class A concrete roof", "runner_up": "CMU + concrete roof"},
  "IMW": {"best": "CMU + concrete roof", "runner_up": "SCIP walls + roof"},
  "DSW": {"best": "ICF (cement-based finish) + concrete roof", "runner_up": "CMU + concrete roof"},
  "CA": {"best": "SCIP walls + roof + WUI package", "runner_up": "CMU + concrete roof (tied with ICF + concrete roof)"},
  "PNW": {"best": "ICF + concrete roof", "runner_up": "ICF + wood-truss roof"},
  "AK": {"best": "Engineered wood, super-insulated (wood roof)", "runner_up": "SCIP walls + roof"},
  "HI": {"best": "CMU + concrete roof", "runner_up": "SCIP walls + roof"},
  "PRVI": {"best": "SCIP walls + roof", "runner_up": "CMU + concrete roof (tied)"}
}

Overlays (applied on top of the state colour; county FIPS given where the zone is a fixed county list):

Wider than the screen. Scroll this block sideways.

[
  {"id": "OV-HVHZ",      "counties_fips": ["12086", "12011"], "layer": "Florida Building Code HVHZ"},
  {"id": "OV-VZONE",     "layer": "FEMA NFHL flood zones V, VE, Coastal A"},
  {"id": "OV-HURRCOAST", "layer": "ASCE 7-22 windborne-debris region (Risk Category II)"},
  {"id": "OV-TORN250",   "layer": "FEMA P-361 safe-room design wind speed = 250 mph"},
  {"id": "OV-NMSZ",      "layer": "USGS NSHM 2023 / FEMA SDC maps, SDC D around New Madrid and Charleston"},
  {"id": "OV-WUI",       "layer": "State WUI / fire hazard severity maps (CA FHSZ, CO CWRC, OR R327)"},
  {"id": "OV-TERMITE",   "layer": "IRC termite infestation probability map, 'very heavy'"},
  {"id": "OV-CLAY",      "layer": "USGS I-1940 swelling clays"},
  {"id": "OV-TSUNAMI",   "layer": "State tsunami inundation maps"}
]

2. Methodology

The map's default picks come from Stage A alone: demonstrated hazard performance. Stage B (cost, labour, code paths) is scored on two separate axes that the reader can weight in with the map's sliders. Each score is for the whole house and rests on demonstrated performance; the evidence depth behind it is shown as a separate badge.

2.1 Stage A: hazard performance

Hazard weighting per state.
- Building-value Expected Annual Loss was pulled for 11 NRI hazards for every state and territory from FEMA's NRI v1.20 service [N-36]. The NRI computes expected loss as exposure × annualized frequency × historic loss ratio [N-34].
- Life-safety hazards were weighted up, because expected-loss figures understate deaths from tornadoes and earthquakes. About 95% of tornadoes are EF0–EF2, but the rest kill [N-55].
- Hazards the NRI does not score were added from primary sources: termites [N-74–N-78], expansive soils [N-70–N-73], climate and energy [N-66, N-69], humidity and decay [MAT-07, N-80], and snow load [N-63, N-64].
- Site design values come from ASCE 7-22 (wind, tornado, snow, seismic via the ASCE Hazard Tool) [N-03, M-24], the USGS 2023 National Seismic Hazard Model [N-01, N-02, M-26], NOAA storm records [M-17, N-53] and IBHS research.
System capability matrix (Section 5.2). Each system was scored against each hazard from its demonstrated performance and engineering (test results, ratings, load-path mechanics), and separately given an evidence-depth badge. The badge never changes the score:
- Extensive data (formerly grade A): a code-prescriptive path plus independent field or lab evidence.
- Moderate data (formerly grade B): an evaluation report or product approval, or an independent lab test.
- Emerging data (formerly grades C and D): a manufacturer-commissioned test, a single study, claims only, or engineering reasoning with no test yet.
Code paths are scored separately. A long or missing code path does not lower the structure score; it lowers the "ease of permitting" score. Thin evidence does not lower it either; it changes the badge. Examples of paths that need engineering or special approval:
- SCIP shear walls above SDC B, unless the specific product's evaluation report covers the SDC (ICC-ES ESR-2435 does; ESR-5623 and ESR-3429 do not);
- IRC SIPs above 155 mph Exposure B or 140 mph Exposure C, or above SDC C [SIP-02];
- IRC R608 concrete above SDC C without engineering [N-04].
Approvals are product-specific. A test or approval for one SCIP, ICF or SIP product is not transferred to another. The order of checks is: specific product → specific evaluation report or approval → code edition → site-specific engineering.

2.2 Stage B: deployability (the price and permitting axes)

Buildability and cost.
- Census framing shares [COST-03, N-83], known manufacturers and installers, and the thin neutral cost literature [COST-05, COST-09–COST-12] set the "best potential price" score. They do not affect the structure score.
- No neutral, current, regional cost study exists for SCIP, ICF or CMU against wood. This report therefore gives no cost-savings figures beyond those quoted from a cited source.
Code adoption, enforcement and insurance.
- Sources: FEMA BCAT [N-41], IBHS Rating the States [M-11], state FORTIFIED mandates and grants [N-44, N-45, COST-16, COST-17, M-12], and insurance stress (California FAIR Plan, TWIA, Treasury FIO) [N-46, N-48–N-50].
- Weak enforcement raises the value of systems that carry their own verification, such as FORTIFIED evaluators, product approvals and special inspection.
- Insurance is scored as a separate, fast-changing layer, not as a structural property. FORTIFIED is easier for insurers to price because it is a third-party-verified standard, not because wood is inherently more insurable than concrete.
Pick and runner-up.
- The best system has the highest structure score: hazard-weighted capability of the whole house, with no evidence adjustment. Cost and availability are not applied.
- The runner-up is the second structure score. Scores within 0.2 points are treated as a tie and go to the system with the stronger demonstrated performance on the region's heaviest-weighted hazard.
- Earlier editions weighed cost, labour and code paths, and then evidence grades, in the pick. The changes are recorded in Section 2.4.
Verification.
- Every negative or "unverified" claim used to rank a system was re-searched against primary sources (ICC-ES, Florida Product Approval, Miami-Dade Product Control, FEMA, IBHS, USGS) before publication.
- The 28 Oct 2026 revision re-searched every negative SCIP claim, and the whole-house revision re-searched them again against the primary documents. Two were wrong and are corrected: "no ICC-ES SCIP report above SDC B" (ESR-2435 reaches SDC F) and "no SCIP has a Miami-Dade NOA" (RSG's 3-D wall panel has one) [SCIP-27, SCIP-28].
- Corrections in the whole-house revision (each re-checked against the primary document):
  - FL30050's ±124 psf and the 9-lb 2x4 at 80 ft/s (54.5 mph, TAS 201) apply to wall panels (PSM 120–240). The earlier text said the approval covers roof slabs at ±124 psf. The slab panels are listed in the approval, but the TER publishes only gravity-load tables for roof panels and no uplift rating [SCIP-08].
  - RSG's NOA 25-0421.03 rests on a 1994 test of a wall panel system; its roof drawing shows a wood-truss roof [SCIP-28].
  - ESR-2435's 1.5 h and 2 h ratings are for walls only and the report cites no ASTM E119 test; ESR-5623 and ESR-3429 explicitly list E119 tests for a 1 h wall and a 1 h floor-roof [SCIP-27, SCIP-04, N-14].
  - GCT's fire report (DrJ 1201-04) was located and confirmed as calculated under IBC 722, not furnace-tested [SCIP-12].
  - No SCIP roof uplift test and no SCIP roof debris-impact test was found after repeated searches.
- Exterior-face revision (30 Oct 2026). Every negative finding behind the new row was re-searched with a second query or source and marked verified or unverified. Verified absences: no NFPA 285 requirement for IRC / Type V houses; no State Fire Marshal (SFM 12-7A-1) listing for an ICF, EIFS or SCIP wall in the 2022, 2024 and 2025 listed-products handbooks; no FEMA assessment-team observation of an ICF house's finish; no SCIP wildfire test; no statistical survival study that separates ICF or foam-backed stucco. Two first drafts were wrong and are corrected: "no ICF finish impact test" (Texas Tech tested vinyl, brick and EIFS on ICF at about 100 mph [EV-08]) and "no field case of ICF foam melting in a wildfire" (one trade-press case, Valley Fire 2015 [EV-21]). Unverified: the live SFM listings database (it blocked access), the ICC-ES AC353 text (paywalled), NFPA 1140 wall provisions and the 2025 FORTIFIED standard. Since the 30 Oct 2026 revision the SCIP roof rows are scored on an engineering basis like the other concrete decks (9), and the missing tests are shown only by an "Engineered / limited formal testing" badge (Section 2.4).
- Termite, insurance and latent-moisture revision (30 Oct 2026). Every negative finding was re-searched with a second query or source and marked verified or unverified. Corrected: the foam-near-grade rule in Florida is FBC Building 2603.8 (statewide, no "very heavy" qualifier) and FBC Residential R318.4; FBC 1503.7 is a downspout-discharge rule and FBC 1816 a soil-treatment rule, neither about foam [TM-01, TM-02, TM-03, TM-05]. Verified absences: no US insurer, rating manual or regulator found that declines or surcharges ICF homes, over termites or anything else; no SCIP termite test; no hygrothermal or field moisture study of SCIP; no independent field survey of mold in ICF homes. Unverified: the exact FBC Residential R318.4 wording (the page quotes the model IRC text), zone assignments for the Northeast and Appalachia (map read only), the US Virgin Islands' zone, the ISO homeowners form text on insect damage, Texas's ICF discount rule, and the 2002 PCA ICF moisture report (cited as reported by ICF Builder).
- Personal testimony (the founder's Ecuador observation in Section 3.3) is reported but not scored.
- Where two research drafts disagreed on a figure, the primary document decided it.

2.3 What this method cannot do

It does not price systems in a given ZIP code. The price axis is a rough regional ordering, not a quote.
Hazard weights per region and the capability scores are editorial judgements from the cited evidence. They are published in full in "How the scoring works" above the report so they can be challenged.
It cannot replace a site check: the ASCE 7 Hazard Tool, the FEMA flood map and the state WUI map.
NRI v1.20 moved inland flooding to a new probabilistic model, so its figures are not comparable with v1.19 "Riverine Flooding" [N-35].

2.4 Scoring for the map

Each system gets three scores from 0 to 10 in each region: structure, best potential price and ease of permitting. The map colours each region by the weighted total; the default weighting is structure 100%, so the map and the region table show structure-only picks. How the scoring works lists every hazard weight and capability score.
Seven systems. ICF and grouted CMU each appear twice, with a wood-truss roof and with a concrete roof deck tied into the walls. Wood frame (wood roof), SCIP (concrete wall-and-roof panels) and SIPs complete the set.
Structure is the sum of hazard weight × capability, with no evidence adjustment. Capability comes from the Section 5.2 matrix, anchored on the dots (strong 9, middle 6, weak 3, none or failed 1) and set by the performance the tests and ratings show. Roof wind performance (uplift and roof-to-wall load path) and the roof's fire and ember path are two rows of their own; a wood-truss roof scores the same whatever the walls, apart from half a point for better strap anchorage in concrete. A concrete roof loses half a point on earthquake for its mass. The termite row (rebuilt on 30 Oct 2026) scores the whole house on two parts: whether the structure, roof framing included, is food, and whether foam or a hollow cavity hides a path from the soil to wood (Section 4.8). The moisture row is the mean of latent moisture (trapped water, drying, condensation, mold), decay and corrosion (Section 4.8). An exterior-face row (new on 30 Oct 2026) scores the skin of the walls: fire, embers and radiant heat on the face, debris and impact on the finish, and repair (Section 3.8).
Evidence depth is a badge. Extensive, Moderate or Emerging data, or "Engineered / limited formal testing" for a row scored by rational analysis of a reinforced-concrete section with no assembly test (the SCIP roof), sits beside each score in the click panel and the capability table. It never changes a total.
Hazard weights. Inside the wind family, the roof gets 40%, debris 25% and wall wind 35%; where wind is not a top threat the old roof weight moves to the roof-fire and ember row, along with 40% of the old fire weight. The exterior-face row gets 3% in every region (hail, everyday impact, repair) plus 15% of the region's hurricane-wind and wildfire weight (wall wind, debris, roof, embers, wall fire), rounded to the half percent: from 4% in Alaska to 13.5% in the Rocky Mountains. The termite row is set by the region's zone on the IRC termite infestation probability map: 12% "very heavy", 7% "moderate to heavy", 3% "slight to moderate", 1% "none to slight" (the earlier, editorial termite weights are dropped). The other weights are scaled down in proportion so each region still sums to 100%. The weights are published in full under "How the scoring works".
Price and permitting for the new systems. Wood-roof variants take the wall system's scores. A concrete roof is engineered under ACI 318, so it loses 1.5 points of price and 2 of permitting (0.5 and 0.5 in Florida, Hawaii and Puerto Rico / USVI, where cast concrete roofs are the trade default).
Changes from the previous edition (28 Oct 2026, structure-first scoring). That edition added 0.5 for an A evidence grade and took 1 off a C grade, and it scored ICF and CMU as walls with a separate roof (ICF best in 16 regions, wood frame 2, grouted CMU 2). Removing the evidence adjustment and scoring the whole house leaves 18 regions with the same wall family ahead, now with a concrete roof (ICF in 15, wood frame in Alaska and the Intermountain West, grouted CMU in Hawaii). Two regions change family. ICF + concrete roof replaces grouted CMU in the Desert Southwest, because the noncombustible roof deck counts in a wildfire region and ICF's flood score is higher. Wood frame replaces ICF in the Pacific Northwest, on the tie rule. The same ICF walls under a wood-truss roof are the runner-up in 12 regions. Weighting price and permitting in restores wood-roofed systems almost everywhere (see the summary).
Section 5.2's wildfire cell for ICF notes that ESR-1269 lists 2 h and 3 h fire-resistance-rated ICF walls, and requires the exterior EPS to be covered with an approved wall covering [M-21]. No ICF wildfire test was found.
Changes in the 30 Oct 2026 revision (exterior face, termites, latent moisture). Against the previous edition, structure-only winners change in five regions, all by less than 0.2: SCIP now leads California and Puerto Rico / USVI (highest total) and the Intermountain West, the Pacific Northwest and Hawaii (tie rule on earthquake). Wood frame keeps Alaska, and ICF + concrete roof keeps the other 14 regions at totals 0.2–0.55 lower than before. Grouted CMU + concrete roof becomes the runner-up in 11 regions, mostly because it scores best on termites and the exterior face. What each change did: the exterior-face row alone moved California, the Intermountain West and Puerto Rico / USVI to SCIP and the Pacific Northwest to ICF under trusses (Section 3.8); the zone-weighted termite row and the latent-moisture part then moved the Pacific Northwest and Hawaii to SCIP and dropped ICF under trusses from runner-up in Texas and the South Atlantic (Section 4.8). ICF is scored with a cement-based exterior finish and termite protection at grade, which the picks now specify; the thin polymer stucco variant is published alongside (Section 3.8).
Correction, 4 Nov 2026 (totals win). The 0.2 tie window was promoting a lower total when its score on the heaviest hazard was higher. Hawaii (CMU + concrete roof 8.05, SCIP 7.87), the Intermountain West (7.12 against 7.10) and the Pacific Northwest (ICF + concrete roof 6.85, SCIP 6.72) now follow the higher total. A tie-break applies only when two totals are equal after rounding to the hundredth. California and Puerto Rico / USVI are unchanged, because SCIP already had the higher total.

3. Testing the case for SCIP: is it best for hurricane, earthquake and fire zones?

SCIP (structural concrete insulated panel) is an EPS foam core between two welded-wire meshes. Truss wires through the foam tie the meshes together, and both faces are finished with shotcrete or mortar. Its distinctive claim is a monolithic shell: walls and roof in one reinforced-concrete system.

Hazard	Verdict	One-line reason
Hurricane	Agree, with conditions	The only HVHZ-approved SCIP, tested to ±124 psf and the 9-lb missile at 80 ft/s as wall panels; slab panels are approved but no report rates a SCIP roof for uplift or roof impact, so the roof is credited on an engineering basis (badge: Engineered / limited formal testing). Scores 7.9 in Florida, behind ICF and grouted CMU with concrete roofs (8.6, 8.4).
Tornado	Not yet	No single-layer SCIP has an ICC 500 / FEMA P-361 tornado-missile result; ICF does.
Earthquake	Partly agree (revised)	An ICC-ES path to SDC C–F now exists (ESR-2435) and the shake-table data are strong, but no real-earthquake record is documented. Leads California and Puerto Rico / USVI on the total, both by under 0.1. It is ahead on earthquake in the Intermountain West, the Pacific Northwest and Hawaii, and behind on the total; the higher total wins.
Termites	Agree	Shotcrete encases the foam on both faces, so there is no exposed EPS to hide galleries from the soil; termite row 8 against 6.5 for ICF with a concrete roof (4.5 under trusses) and 8.5 for grouted CMU. No SCIP termite test exists (Emerging badge) (Section 4.8).
Fire	Partly agree	Walls rated 1–2 h (1 h E119-tested in two ICC-ES reports; 1.5–2 h in ESR-2435 with no test cited); a concrete roof removes attic vents; but embers, vents, eaves and spacing decide survival, and there is no SCIP wildfire test.
Exterior face	Agree	Reinforced shotcrete covers the foam on both faces, so no EPS is exposed to embers, radiant heat or debris; it scores 8 against 6.5 for ICF with a cement-based finish (4 with thin polymer stucco) and 8.5 for grouted CMU. No SCIP face has a wildfire or hail test (Section 3.8).

3.1 Hurricane: agree, with conditions

For:

The approval exists. Gulf Concrete Technology LLC (GCT, Long Beach, MS) holds Florida Product Approval FL30050-R4 (approved 03/27/2026, FBC 2023). It is approved for use inside and outside the HVHZ, Impact Resistant: Yes, design pressure +124/−124 psf [SCIP-06].
The certification behind it is DrJ TER 1202-12 [SCIP-08 p25]:
- TAS 201: "PSM 120 wall panels resisted the impact of the 9-lb 2 x 4 missile propelled at 80 ft/s (54.5 mph) without damage to the backside."
- TAS 202/203: static and cyclic tests at 124 psf.
- 80 ft/s is the FBC §1626 speed for Risk Category IV; ordinary houses are tested at 50 ft/s [LF-13].
The roof is part of the product line, but not of the tests. The approval's model line is "PSM Series Wall Panels, PSM Series Slabs, PSG Series Slabs and SMP 80 Slab", and the TER title says "for Use as Floors, Walls, and Roofs" [SCIP-06, SCIP-08]. But every HVHZ statement in the TER's section 6.12 says "PSM wall panels": the ±124 psf static and cyclic result (PSM 120–240) and the missile result (PSM 120). The roof tables are gravity-load capacities (allowable load limited to 200 psf), and no section gives a negative (uplift) pressure for a roof or slab panel. The installation instructions detail the wall-to-roof connections, with no wind data [SCIP-11]. An approved, insulated, reinforced-concrete roof deck is still a real differentiator over trusses, since Florida's mitigation form treats a "reinforced concrete roof deck" as a separate answer [OPN-02], but the uplift performance of the SCIP roof is engineering, not a published test. The Florida Building Code accepts that basis for uplift: any system "shall admit of a rational analysis in accordance with well-established principles of mechanics" (§1616.1.1), HVHZ wind loads are designed to ASCE 7 (§1620), and a test is required only where capacity "cannot be calculated by rational analysis" (§1625.3) [N-106]. The scores therefore credit an engineer-sealed SCIP roof like the other concrete decks and show the missing test in the badge.
The roof is where houses fail.
- FEMA's Hurricane Michael team: "the most commonly observed damaged elements of the building envelope … were roof coverings, soffits, and exterior wall coverings", with roof-covering performance on post-FBC houses "generally very poor" [FM-08]. In Ian, "the most common building envelope damage observed was roof covering failure" [FM-11]. In Irma, rafters separated from connectors on a Cudjoe Key house and the roof lifted away, and the team noted that "once wind enters a building, failures can progress to other components and connections along numerous load paths" [FM-07].
- IBHS damage thresholds in Hurricane Michael: roof cover above 110 mph, decking above 130 mph, structure above 140 mph, collapse above 150 mph [FTD-19].
- A sealed deck cut water entry by at least 95% [FTD-16].
- A concrete roof deck removes those failure modes; that is why a concrete roof is scored as its own variant of ICF and CMU, not only as SCIP's feature (Section 3.7).
Concrete-walled homes fare better in wind. MIT CSHub's re-analysis of StEER data found wood-frame wind damage ratios in Irma "roughly twice those of reinforced concrete masonry homes." Caveat: the concrete samples were small and age was not controlled [MAT-02].

Against, or not proven:

Impact resistance is not unique to SCIP, and its roof is not automatically deemed to comply. FBC 2023 §1626.4 deems compliant, without a missile test, "exterior reinforced concrete elements constructed of solid normal weight concrete (no voids), designed in accordance with Chapter 19 … and having a minimum 2 inches (51 mm) thickness", and 8-in. CMU built to Chapter 21 [N-106, N-96]. The conditions are exterior, reinforced, solid normal-weight concrete with no voids, designed under Chapter 19 (whose HVHZ sections 1919–1929 are reserved, so ACI 318 applies) and at least 2 in. thick; no minimum reinforcement or cover is stated, though ACI 318 cover still applies. A SCIP roof's 2-in. exterior wythe sits exactly at the minimum, and whether a foam-core panel counts as "solid … (no voids)" is the building official's call: we found no Florida Building Commission ruling either way. Section 1626 has no rational-analysis substitute for the impact test, so without §1626.4 the routes are a TAS 201/203 test or local alternative-methods approval (§104.11). That is why the SCIP roof's debris credit is labelled engineered, not tested. ICF and grouted CMU stop far heavier missiles [ICF-07, CMU-04].
There is no independent hurricane field record. We found no FEMA, NIST, IBHS or StEER report on a SCIP house in a hurricane.
- Accounts of SCIP Habitat for Humanity houses in Liberty City (and one house in Homestead) surviving Hurricane Andrew trace back to a 1992 newspaper article.
- They are repeated in a FEMA-styled case study hosted on a vendor's website [M-27]. That document is an unpublished draft with editorial notes still in it, and it does not appear in the published FEMA Hurricane Ike report.
- We treat them as anecdotes, not evidence.
Modern-code wood frame already performs well structurally.
- In Michael, structural failures of post-FBC houses were "rare" [FM-08].
- IBHS says the modern FBC "now offers nearly the same protection as the IBHS FORTIFIED Gold™ standard across a large portion of" Florida [FTD-21].
Surge destroys anything that is not elevated [FM-06]. SCIP does not change the V-zone rules [N-37].
There is one Florida-approved HVHZ wall-and-roof supplier.
- FL46983, held by Fortified Structural Solutions LLC (FSS), is not approved for HVHZ use [SCIP-01]. FSS is a company name unrelated to the IBHS FORTIFIED programme. The product was formerly marketed as Concrewall USA and is evaluated in ICC-ES ESR-5623.
- Re-checked 28 Oct 2026 on floridabuilding.org: FL46983-R0 is Approved (Code Version 2023, approved 20 Jan 2025, evaluation entity ICC-ES), "Approved for use in HVHZ: No", "outside HVHZ: Yes", "Impact Resistant: Yes", "Design Pressure: N/A" [SCIP-01]. ESR-5623 itself lists no impact or wind test, and the words "ASCE 7" and "wind" do not appear in its text, so the FSS deck's statement that ESR-5623 "allows use in ASCE 7 wind-designed structures" [SCIP-44] could not be confirmed [SCIP-04]. R1 (name changed to Fortified Structural Solutions) is in Re-Apply status.
- Schnell's ESR-3429 likewise says HVHZ compliance "has not been evaluated" [SCIP-04, N-14].
- Emmedue's DrJ TER is past its renewal date [SCIP-17].
- GCT, FSS, Schnell and Emmedue hold no Miami-Dade NOA [SCIP-24]. One 3-D panel does: Re-Structure Group's RSG 3-D system, NOA 25-0421.03 (HVHZ, large and small missile, expires 2030). Its only listed test dates from 1994, and the applicant listing describes a wall panel system [SCIP-28].
GCT's fire ratings are calculated. Its hourly ratings come from IBC 722, not a furnace test [SCIP-12]. Section 3.4 gives the tested figures for other SCIP products.
Cost is unknown. The only neutral data point is a 1994 HUD case study, in which welded-wire sandwich-panel walls cost 63% more than wood and took 68% more time [COST-09]. There is no current neutral figure.

Verdict: Among concrete shells, a concrete roof deck is what separates the best Florida houses from the rest, and SCIP is one way to get one. On the whole Florida hazard mix, ICF with a cast concrete roof scores highest (8.3), on flood, the tornado missile and durability. Grouted CMU with a concrete roof is 0.06 behind (8.3), ahead on termites and the exterior face. SCIP scores 7.9, third: its roof is credited like the other concrete decks on an engineering basis, and the lack of a published roof uplift or impact test shows in its badge, not its score. In the HVHZ, where wind and the roof weigh most, the gap narrows (Section 1.3). Elsewhere on the hurricane coast ICF also leads, because tornadoes overlap it and ICF and fully grouted CMU are the walls with tornado-missile results.

3.2 Tornado: not yet

No single-layer SCIP has a published ICC 500 / FEMA P-361 tornado-missile result (15-lb 2x4 at 100 mph). Re-checked 28 Oct 2026; RSG's Miami-Dade NOA covers the hurricane large missile only [SCIP-28].
- GCT's TER says its panels were "tested in accordance with TAS 201 and ICC 500 Section 804" but reports only the 9-lb, 54.5-mph result [SCIP-08].
FSS's 2008 Texas Tech missile claim does not change this. A manufacturer presentation says FSS panels were missile-tested at Texas Tech in 2008 with "debris impacts up to 90 mph" [SCIP-44]. It gives no missile mass, standard, panel thickness or report, and we found no Texas Tech report for it. Read exactly: if the missile was the 15-lb 2x4, 90 mph is below the 100-mph vertical-surface criterion of FEMA P-361 / ICC 500 [N-55] (and above the 67-mph horizontal one); if it was the 9-lb 2x4, 90 mph is well above the 54.5-mph (80 ft/s) hurricane-missile speed used in TAS 201 / ASTM E1996 for the large missile [SCIP-08]. Which one is unknown, so it is scored as no result. The deck's "equivalent to approximately 225 mph hurricane conditions" has no stated method and is not used.
The only tornado-speed pass found is Emmedue's 2005 Texas Tech test of a double panel with a 100-mm cast concrete core (PDME100). The test was manufacturer-commissioned [SCIP-26].
- Single panels (PSM 80) passed the 9-lb missile at 34 mph and the 15-lb missile at 66 mph, and were not tested at 100 mph. The double panel passed the 15-lb missile at 100–103 mph, which penetrated 2¾ in. and stopped at the cast core [SCIP-26].
ICF has independent evidence.
- FEMA P-361 accepts ICF with a uniform core ≥4 in. and 6-in. reinforced concrete [N-55].
- Texas Tech shot 6-in. ICF cores at up to 119.9 mph without core damage [ICF-07].
- Fox Blocks also claims an ICC 500 pass, without published test details [N-97].
Whatever the wall, sell the house as a durable house plus a safe room, never as "tornado-proof."
- FEMA requires FEMA-funded residential safe rooms to be designed for 250-mph tornado loads regardless of location [M-05].
- ASCE 7-22's new tornado chapter (Chapter 32, a result of NIST's Joplin investigation) does not turn an ordinary house into a shelter [M-06, M-08, M-09].

3.3 Earthquake: partly agree (revised 28 Oct 2026)

This verdict changed in this revision. The earlier text said both ICC-ES SCIP reports stop at SDC B. That was true of the two reports we had checked, but a third current ICC-ES report reaches SDC F. The code path and the test record are now stronger than the field record, so they are reported separately below.

Code-approval status (what a US permit needs today):

ICC-ES ESR-2435 (Re-Structure Group, "RSG 3-D Structural Panel (3-D/EVG Panels)"; reissued April 2026, revised October 2026, renewal April 2027) covers walls in SDC C, D, E and F. They "must be designed in accordance with Chapter 18 of ACI 318-19 … as Special Structural Walls … conforming to limitations prescribed in ASCE/SEI 7 for special reinforced concrete shear walls" (§4.1.5, §5.4). In SDC A–B it sets R = 3.5, Cd = 3.5, Ω0 = 3.0. It carries California and City of Los Angeles supplements, requires an engineered design, and requires special inspection under IBC 1705.3 and §1908 [SCIP-27].
ESR-5623 (FSS) and ESR-3429 (Schnell) still limit SCIP shear walls to SDC A and B (§5.8 in both) [SCIP-04, N-14]. Approvals are product-specific: their panels cannot borrow ESR-2435's scope.
GCT ("certified for all seismic zones").
- GCT's seismic credential is DrJ TER 1202-12, a Technical Evaluation Report from DrJ Engineering, an ISO/IEC 17065 certification body accredited by ANAB. It is not an ICC-ES ESR or an IAPMO UES report. We found no GCT listing at ICC-ES, IAPMO UES, LADBS or Intertek [SCIP-08].
- The TER states that its seismic table applies to "all seismic design categories". It gives special-reinforced-concrete shear-wall coefficients (R = 5, Ω0 = 2.5, Cd = 5), with no height limit in SDC B–C, 160 ft in D–E and 100 ft in F. It covers the 2018–2024 IBC and IRC [SCIP-08 p12, p26, p28].
- The basis is ASTM D7989 equivalency plus E72/E564/E2126 testing. The TER calls the D7989 data "intellectual property and/or trade secrets", and its substantiating-data list names no cyclic test report [SCIP-08 p28, p30].
- GCT's own marketing describes the R factor as "a controlled simulation by Carmelo carried by DrJ Engineering" [SCIP-34].
- We did not find the words "all seismic zones" on GCT's or its parent company's pages. The nearest wording is "seismic resistance designs for any structure" [SCIP-34] and "R of 5 … in seismic design category zones D, E & F" [SCIP-33].
- So the claim holds in substance, but only under GCT's own DrJ listing. A building official accepts a DrJ TER case by case, as a research report under the alternative-materials section, and may require peer review [N-101].
SCIP is not a named system in ASCE 7 Table 12.2-1. A SCIP product without an SDC-qualified report would need the ASCE 7-22 §12.2.1.1 route: test data, approval by the authority having jurisdiction, and an independent structural design review. That is impractical for a single house.

Demonstrated performance (tests and field results):

Full-scale shake table (Bologna and Eucentre, Pavia, 2011). A 3-storey building with sprayed-concrete sandwich walls was shaken from 0.05 g to 1.2 g [N-11, SCIP-32].
- "Up to the 1g seismic test … the structure did not show any damage."
- Only hairline cracks followed. The fundamental frequency fell from 10 to 8.2 Hz, and the structure survived the final 1.2 g run.
- The journal version reports stiffness and strength "quite superior" to what earlier cyclic tests predicted [SCIP-32].
- The same authors note that these systems are "widely used … in non seismic areas or in areas of low seismicity" [N-11].
- The panels were Nidyon NYSP, made by Nidyon Costruzioni (Rimini), which funded the research. The test is evidence for this class of panel, not for a particular brand such as Emmedue [SCIP-41].

Eucentre (Pavia) shaking table, 7 December 2011. The full-scale 3-storey sandwich-wall building from the test reported in [N-11] and [SCIP-32]. The input was the 1979 Montenegro (Ulcinj) record scaled up to 1.2 g; there was no damage up to the 1 g run, and hairline cracks only after 1.2 g. The video was posted by the YouTube channel of Salvador Ivorra; S. Ivorra-Chorro (University of Alicante) is a co-author of [N-11]. The paper is the source; the video only illustrates it [VID-01].

Full-scale forced-vibration test of an Emmedue house (University of Perugia, published 2003). The RITAM laboratory built a two-storey Emmedue prototype, 3.45 × 4.2 m in plan, and shook it with a vibrodyne (a mechanical shaker fixed to the structure) rather than a shaking table [SCIP-42].
- The abstract reports response accelerations of 10 m/s² (about 1 g) measured on the structure, which the authors say "have clearly shown the capacity of resistance against dynamic actions of this building system".
- This is a structural response, not a ground acceleration, so it cannot be compared directly with the 1.2 g table input above. Only the abstract is online; it reports no damage observations.
Other 3-D panel shake-table videos (supporting links only; none has a published report giving its peak acceleration):
- Emmedue's two-storey model on the ENEA Casaccia (Rome) shaking table (programme presented at ENEA on 27 Nov 2008; video April 2009). The only programme document found covers the static and cyclic wall and floor tests; we found no published shake-table results [VID-02, SCIP-39]. India's BMTPC lists an ENEA document, "Analysis and validation of the static and seismic behavior of EMMEDUE buildings", among the papers it assessed, but does not give its results [SCIP-43].
- A full-scale two-storey M2 (Emmedue) module on the PUCP shaking table in Lima: PUCP report INF-LE-350-08, March 2009, commissioned by the Peruvian licensee. The report is not online. A secondary summary says the module passed four phases up to a "catastrophic" level, scaled against Peru's May 1970 earthquake, with small horizontal surface cracks and no diagonal cracking. It gives no peak acceleration [VID-03, SCIP-40]. BMTPC's 2014 certificate, which assessed the report, records that "the module remained quite stable" and "passed the seismic test" [SCIP-43].
- Emmedue's own compilation of seismic tests on two-storey structures [VID-04].
- A clip labelled as a seismic test of Emmedue panels at the University of Perugia, posted by a third-party channel around 2009. We could not tie it to [SCIP-42] or to any other report [VID-05].
Scale shake table (Iran, 2009). A 1:2.35-scale 4-storey 3-D panel building had micro-cracks at 0.3 g and cracking at openings and corners above 1.0 g (model-scale accelerations). The authors concluded "certain life safety and good overall stability", but warned of "local collapse" at unstiffened corners [N-12].
Cyclic wall tests.
- Emmedue walls in the European assessment reached ductility 2.8–6.6 [SCIP-18].
- Three full-scale SCIP walls tested at Idaho State University (MS thesis) reached ultimate moment at 0.87% drift, with a calculated overstrength of 2.5 [SCIP-30].
FEMA P-695 analysis (Univ. at Buffalo). R = 3.5, Cd = 3.5 and Ω0 = 3.0 pass for one- and two-storey 3-D panel buildings in SDC D. Taller archetypes failed, and the test data behind the models were rated only "medium" quality [SCIP-29]. These are the coefficients ESR-2435 uses in SDC A–B.
Mass. GCT walls weigh about 35 psf [SCIP-08], against the code's 85 psf for 6-in. concrete and 15 psf for a wood wall [N-04]. Seismic force scales with mass.
Real earthquakes: no engineering record found.
- We searched the EERI, EEFIT and Build Change reconnaissance reports on the M7.8 Ecuador earthquake of 16 April 2016 in full text. None mentions SCIP, polystyrene-core or 3-D panel houses, or the Ecuadorian brand Hormi2 [SCIP-36, SCIP-37, SCIP-38].
- Searches for Chile 2010, Mexico 2017, Turkey 2023, L'Aquila 2009, Emilia 2012, central Italy 2016, El Salvador 2001, Pisco 2007, Haiti 2010, Nepal 2015 and Albania 2019 also found no engineering or peer-reviewed account. Claims that SCIP houses "survived" these events come from manufacturers and licensees.
- This means no one has documented it. It does not mean the houses failed.
"Widely used in high-seismic zones."
- Partly documented. Peru approves Emmedue/M2 for up to 2 storeys and EVG-3D for up to 5 storeys in its highest seismic zone, on firm soil (secondary source: a 2024 thesis citing the ministerial resolutions) [SCIP-31].
- The European assessment is written for "dissipative structures in seismic areas" [SCIP-18].
- We found no production or usage figures by seismic zone, so "widely" cannot be quantified.

Founder's field observation (personal account, not a survey or engineering report)
BuiltStrong's founder was in Ecuador after the M7.8 earthquake of 16 April 2016. He saw SCIP homes that came through with no cracks he could see, while many other buildings around them had failed. That experience is why he started building with SCIP.
What it can and cannot show: it is one person's observation of the houses he saw. It does not record how many SCIP houses were in the shaken area, how strong the shaking was at each, which panel brand or detailing they used, or how comparable houses of other systems fared. We could not find any engineering reconnaissance report that recorded SCIP houses in this earthquake (see above), so the observation cannot yet be checked against independent data. It is consistent with the shake-table results, but it does not enter the scoring. Addresses, photographs, builder names and panel brands would let it be documented properly.

What a builder must do to permit SCIP in SDC D–E:

Use a product whose evaluation report covers the site's SDC. Today that means ICC-ES ESR-2435, or GCT's DrJ TER 1202-12 where the building official accepts it. ESR-5623 and ESR-3429 products stop at SDC B [SCIP-27, SCIP-08, SCIP-04, N-14].
Have a registered design professional engineer the house. For IRC houses this means engineered design under IRC R301.1.3. The walls are designed as ASCE 7 special reinforced-concrete shear walls with ACI 318 Chapter 18 detailing. Boundary and opening reinforcement may go beyond the panel mesh; that is the engineer's call, not something the report settles [SCIP-27].
Apply for approval as an alternative material. This is 2024 IBC 104.2.3 / IRC R104.2.2, or 2021 IBC 104.11 / IRC R104.11. Submit the evaluation report and calculations, and expect a possible peer-review requirement (2024 IBC 104.2.3.7) [N-101, N-102].
Budget for special inspection of reinforcement placement and shotcrete (IBC 1705.3, §1908) and for a Class A roof-assembly justification where a SCIP roof is used (ESR-2435 §5.8) [SCIP-27].
In California, check the supplement edition. ESR-2435's supplements cite the 2022 CBC/CRC, while the 2025 codes took effect on January 1, 2026. Confirm acceptance with the building department early.

Verdict: The code path now exists, and the lab evidence is good. The field record is missing, so the Moderate data badge applies, not a lower score: SCIP scores 6.5 for earthquake, above ICF and CMU houses (5.5–6) because it weighs about 35 psf, and below wood frame (9). On structure alone it leads California and Puerto Rico / USVI, both by under 0.1. It is the runner-up in the Intermountain West, Hawaii and Alaska, and third in the Pacific Northwest, Florida and the Desert Southwest. An earlier 0.2 tie window had let its earthquake score overturn the higher totals in the Intermountain West, the Pacific Northwest and Hawaii; totals now win. US supply is thin, which affects the price score, not the structure score. Wood frame still leads Alaska clearly.

What would move SCIP to best in an earthquake region:

a documented post-earthquake survey of SCIP houses;
a US full-scale test of a code-detailed SCIP house;
local plants and crews.

3.4 Fire (wildfire): partly agree

Fire-resistance ratings: what is tested and what is calculated.

Product	Assembly	Rating	Basis	Source
FSS (ESR-5623) / Schnell (ESR-3429)	Wall: 3.15 in. EPS + 1¾ in. concrete each face (6.75 in.); floor/roof 10.65 in.	1 h	Furnace-tested, ASTM E119	[SCIP-04, N-14]
RSG 3-D / EVG (ESR-2435)	Wall, concrete faces 1.5 in. / 2 in. each side	1.5 h (carbonate) or 1 h (siliceous) / 2 h	ICC-ES rating; the report cites no ASTM E119 test (its evidence list is AC15 data only) and does not say whether tested or calculated. Roofs and floors are left to the designer	[SCIP-27]
Emmedue (ETA 20/0976)	Wall, 40 mm shotcrete + 80 mm EPS + 40 mm shotcrete, loaded	REI 90 / RE 120	Furnace-tested, EN 1365-1	[SCIP-18]
Emmedue PSM80 (manufacturer)	Loaded wall	2 h	Manufacturer announcement of a test at a Saudi laboratory. No standard or report number published	[SCIP-35]
Emmedue (DrJ TER 2110-01, legacy)	Wall, average 1.38 in. mortar each face	2 h "when tested according to ASTM E119"	Lab and report not named; report no longer maintained	[SCIP-17]
GCT (DrJ TER 1201-04)	Wall: ¾ in. under mesh + 0.5 / 1.25 / 2 / 2.5 in. mortar per face	1 / 2 / 3 / 4 h	Calculated under IBC 722, not furnace-tested	[SCIP-12]
GCT (marketing)	"Standard PSM80" wall	1 h	Pamphlet says "tested … according to the ASTM E-119"; no report cited, and it conflicts with the TER	[SCIP-33]

FSS's demo-day deck repeats the ESR-5623 fire claims; it adds no new test. It says the EPS core meets IBC 2603 and that complete wall and roof assemblies achieved 1 hour in ASTM E119 [SCIP-44]. ESR-5623 does list the tested 1-h wall and 1-h floor-roof (restrained) [SCIP-04], so that part checks out. The deck gives no lab, date or report for the furnace tests either.
"Usually 2+ hours, rising with concrete thickness" is half right.
- Ratings do rise with face thickness: ESR-2435 goes from 1–1.5 h at 1.5 in. to 2 h at 2 in. [SCIP-27].
- Published furnace tests of SCIP walls sit at 1–2 hours [SCIP-04, SCIP-18]. Every 3–4-hour figure is calculated [SCIP-12].
- A typical ¾-in. mortar cover over the mesh is below what GCT's own fire report needs for 1 hour [SCIP-12].
- Specify the face thickness for the rating you need.

For (wildfire):

The faces are noncombustible concrete.
A concrete roof slab removes the vented attic, one of the main ember entry paths [N-18, M-13]. That this applies to a SCIP roof is our inference; it has not been tested.
The 1–2-hour ratings exceed what a WUI exterior wall needs.

Against (wildfire):

In California structure-loss data, siding and roof material explained far less than eaves, vents and windows [N-18].
In the 2025 LA fires, noncombustible siding was on 76% (Palisades) and 89% (Eaton) of surveyed homes. Survival tracked spacing, Zone 0 fuel and the number of hardened features [N-22].
In the Camp Fire, most damage to surviving homes came from radiant heat from neighbouring burning structures [N-19].
No SCIP wildfire or façade test exists (ASTM E2707, SFM 12-7A-1, NFPA 285 or EN 13501 façade); re-checked 28 Oct 2026.
A SCIP roof still needs a justified Class A roof assembly (ESR-2435 §5.8) [SCIP-27]. In Florida an exposed concrete roof deck counts as Class A without testing (FBC §1516.2.1), and no separate thermal barrier is needed over foam covered on each face by at least 1 in. of concrete (§2603.4.1.1) [N-106]. Hourly roof ratings still need an ASTM E119 test (1 h in ESR-5623 and ESR-3429) or an engineering analysis the official accepts; a 2-in. wythe alone is below the 3.5 in. (siliceous aggregate) the solid-concrete tables need for 1 h.

Verdict: SCIP is a good fire shell, and with a concrete roof slab it removes more ember paths than a wood-roofed house (roof fire and ember row: 9 against 5, scored like the other concrete decks on an engineering basis, with an "Engineered / limited formal testing" badge because no SCIP roof fire or wildfire test exists). Its wall scores 7 against 9 for ICF and grouted CMU, because its rated 1–2 h is shorter than ICF's 2–3 h and CMU's 2 h or more, and there is no SCIP wildfire test (an Emerging data badge). On its face, though, it beats ICF: shotcrete covers the foam, so it scores 8 on the exterior-face row against 6.5 for ICF with a cement-based finish and 4 with thin polymer stucco (Section 3.8). Fire alone does not make it the pick in a wildfire region; its wins in California and the Intermountain West come from earthquake, with the exterior face and the termite row closing the gap. The house-level package (vents, eaves, glazing, Zone 0, spacing) still matters more than the wall.

3.5 Bottom line on the SCIP lean

SCIP is one of the most promising multi-hazard residential systems. Its strengths are wind, debris, fire, an integrated concrete wall-and-roof shell, and foam that is encased in concrete on both faces, so it has no exposed foam for fire, debris or termites. On structure alone, scored on the evidence in this report:

Best (two regions, both close calls on the total): California (7.11 against 7.04 for CMU + concrete roof) and Puerto Rico / USVI (7.76 against 7.75). It is behind the higher total in the Intermountain West (7.10 against 7.12), the Pacific Northwest (6.72 against 6.85) and Hawaii (7.87 against 8.05). It scores 6.5 on earthquake against 5.5 for the concrete-roofed houses, and that breaks only an exact tie. Its low weight, concrete roof slab and shotcrete faces carry it, and in the "very heavy" termite zones of California, Hawaii and Puerto Rico its encased foam helps (termite row 8 against 6.5 for ICF). The HVHZ overlay lists it as one of three concrete-shell options (Section 1.3).
Runner-up: Alaska (6.61; wood frame leads at 7.61).
Third: Florida and the Desert Southwest.
Fourth: the South Atlantic, the Northeast and Mid-Atlantic, Northern New England and the Rocky Mountains.
Fifth: the Gulf Coast, Texas, Appalachia and tornado country, behind ICF and CMU with concrete roofs and ICF with a wood roof. Tornado country counts against it most, because no single-layer SCIP has a 100-mph tornado-missile result (score 4).

The gap is a score gap on specific hazards, not only a data gap: the 100-mph missile, flood and wall fire rating are lower than ICF's. The roof is not part of the gap: it scores like the other concrete decks on an engineering basis, and its lack of a published uplift or impact test shows only in the badge ("Engineered / limited formal testing"). Where the data is thin the badge says so and the number is engineering judgement from the nearest tested assembly or from rational analysis. What would move SCIP up: a single-layer ICC 500 tornado-missile result, a documented post-hurricane or post-earthquake survey of SCIP houses, a wildfire or façade test, and a flood-damage-resistance listing. A roof uplift and roof-impact test would deepen the evidence badge, and so would an ASTM E2707 / SFM 12-7A-1 test of its exterior face. Its two wins sit within 0.1 of the next system, so a small change in any weight can move them. In an engineered design under an evaluation report that covers the site's SDC, it remains a credible option in every concrete-shell region.

3.6 Earthquake and wildfire regions: code approval versus demonstrated performance

"Code-approval status" is what a US permit needs today. "Demonstrated performance" is what tests and real events show. Changes made in this revision are marked "Before".

Region	Code-approval status (what a US permit needs today)	Demonstrated performance (tests and field results)	SCIP placement (structure only, of 7 systems)
CA	SDC D–E across most of the state [N-03]. Use ESR-2435 (California and City of LA supplements) or an accepted DrJ TER. Engineered special-wall design, alternative-material approval, special inspection, Class A roof justification; 2025 California WUI Code in fire zones [SCIP-27, N-25]	Shake table to 1.2 g, hairline cracks only [N-11]; no real-earthquake record; no wildfire test; furnace-tested walls 1–2 h; shotcrete faces leave no foam exposed	1st, the highest total by 0.07 (CMU and ICF + concrete roof tied behind). Before: 2nd
PNW	SDC D–E (Cascadia). Same route as CA, without the CA supplements; WUI detailing voluntary in WA [N-30]	As CA; rain-driven moisture is the chronic hazard, and no SCIP moisture study exists	3rd (ICF + concrete roof 0.13 higher). Before the tie-window bug was removed: 1st on that rule
AK	SDC D–E; no statewide residential code, local codes in Anchorage and other cities. Same route as CA	As CA; IECC 7–8 cold exposes truss-wire bridging and calculated R-values [N-68]	2nd (wood frame best). Before: 3rd
IMW	Wasatch Front SDC D–E; much of NV, ID, MT C–D [N-03]. Same route as CA	As CA; wildfire risk to homes very high in NV, ID, UT	2nd (CMU + concrete roof 0.02 higher). Before the tie-window bug was removed: 1st on that rule
DSW	Mostly SDC B–C, with D pockets [N-03]. ESR-2435, ESR-5623 and ESR-3429 all usable in A–B; C and above needs ESR-2435 or an accepted TER	As CA; hot-dry climate favours an insulated concrete roof (our inference) [N-69]	3rd (ICF + concrete roof best) on termites as well as fire. Before: 4th
ROCK	Mostly SDC B–C [N-03]. Colorado Wildfire Resiliency Code in the WUI [N-29]	No hail or wildfire test; cold climate exposes wire bridging	4th, 0.02 behind ICF under trusses. Before: 4th
SATL (Charleston SDC D overlay)	ESR-2435 route in the overlay, any SCIP report outside it	Tornado missile untested for single-layer SCIP [SCIP-26]	4th (ICF + concrete roof best). Before: 5th
MIDS (New Madrid SDC D overlay)	As SATL	As SATL; tornado is the governing life-safety hazard	5th
HI	Hilo and the Big Island SDC D–E [N-02, N-03]. ESR-2435 route; salt-spray durability outside every SCIP report's scope [N-14]	As CA; ~1-in. cover thin for salt spray [SCIP-11]; no termite test in a "very heavy" Formosan zone	2nd (CMU + concrete roof 0.18 higher). Before the tie-window bug was removed: 1st on that rule
PRVI	SDC D across most of Puerto Rico [N-03]. Puerto Rico Building Code 2018 is IBC-based; GCT's TER lists it [SCIP-08]	As CA; no SCIP record from Maria (2017) or the 2020 earthquakes found	1st, 0.01 ahead of CMU + concrete roof (ICF + concrete roof 0.16 behind). Before: 2nd

3.7 The whole house: walls, roof and the connection between them

Most ICF, CMU and wood homes are built with wood-truss roofs. The scoring therefore treats the roof as part of the structure, with ICF and CMU each scored twice.

What the field evidence says about roofs

Hurricanes. FEMA's assessment teams report roof coverings as the most commonly damaged envelope element in Michael (2018) and Ian (2022) [FM-08, FM-11]. The Irma team documented roof-to-wall connection failures where rafters separated from their connectors and the roof lifted away [FM-07]. Once wind breaches the roof, "failures can progress to other components and connections along numerous load paths" [FM-07]. The Michael team also warned that unreinforced masonry walls can topple in high wind [FM-08], so reinforcement and grouting, not "block", are what count.
IBHS. A sealed roof deck is the backup when the covering is lost, and cut water entry by as much as 95% in IBHS tests [FTD-16]. FORTIFIED homes had loss frequency 55–74% lower in Hurricane Sally [FTD-09, M-10]. We found no study that isolates the roof-to-wall connection from the other FORTIFIED requirements.
Wildfire. IBHS: "Wind-blown embers are the principal cause of building ignitions during wildfires", and vents are among the most vulnerable points of a structure [M-13]. FEMA's Marshall Fire team: unsealed vents "provide a pathway for embers to enter the house, resulting in possible ignition inside the structure" [N-103]. A noncombustible, unvented roof deck removes that path; the 2021 model IBC and IRC treat an exposed concrete roof deck as a Class A roof assembly [N-104, N-105]. The 2024 text was not verified from a primary source.

How the scores use it

A wood-truss roof scores the same on roof wind performance (6–6.5) and on the roof fire and ember path (5) under ICF, CMU or wood walls. The walls only improve strap anchorage.
A concrete roof deck tied into concrete walls scores 9 on both rows. It costs half a point of earthquake score for the extra mass, and money and permitting ease (Section 2.4).
SCIP's roof scores 9 on both rows, like the other concrete decks. It is a continuous steel-reinforced shotcrete envelope (typically a 2-in. exterior and 1.5-in. interior wythe, monolithic with the walls), and its uplift, roof-impact and roof-fire performance is credited on an engineering basis: a reinforced-concrete section designed by rational analysis and sealed by a registered engineer. No SCIP roof uplift or roof-impact test has been published, so those rows carry the "Engineered / limited formal testing" badge; the missing test lowers the badge, not the score. The credit assumes the roof is actually engineered and built that way, with an approved roof covering, and a Class A justification is still needed (ESR-2435 §5.8) [SCIP-08, SCIP-27].

Where a wood-roofed system still wins on structure alone. Alaska only. Earthquake carries 71% of the weight there, wood frame scores 9 for earthquake against 5.5–6.5 for the concrete houses, wind barely counts and termites are "none to slight", so the margin is large (7.6 against 6.6). In the Pacific Northwest ICF under a wood-truss roof is within 0.1 of the leaders and scores higher on earthquake than ICF with a concrete roof, but SCIP scores higher still (6.5) and takes the tie. Wood frame no longer wins the Pacific Northwest or the Intermountain West: it sits 0.3–0.4 behind the leaders there, outside the tie window (Sections 3.8 and 4.8). A wood-truss roof also loses on termites wherever they weigh: it is the wood that termites reach through ICF's foam. The click panel states this when a wood-roofed system leads.

3.8 The exterior face: exposed foam, finishes and repair (new, 30 Oct 2026)

Why this row exists. The model credited ICF's concrete core for fire, wind and debris, but said nothing about what sits outside it: about 2.5 in. of EPS under a finish. SCIP has 1.25–2 in. of reinforced shotcrete outside its foam, grouted CMU usually has stucco directly on the block, and wood frame and SIPs have cladding over combustible sheathing. The exterior-face row scores that skin for every system, separately from the wall's fire rating (the "Fire and wildfire (wall)" row) and its structural debris resistance (the debris and tornado rows).

What it scores. Three parts, scored 1–10 and averaged, rounded to the half point:

Fire, embers and radiant heat on the face: does the face ignite, melt back, pool-burn or expose foam?
Debris and impact on the finish: is the finish holed, stripped or cracked, and does water get in, even where the structure holds?
Repair after an event: how local, conventional and inspectable is the fix?

What the codes say.

For ordinary houses the codes treat ICF fairly. NFPA 285 applies only to Type I–IV buildings; Type V and IRC houses go to the foam-plastic rules (IBC 2603.5; IRC R316, renumbered R303 in 2024) [EV-03, EV-02]. The IRC's only rule on exterior foam is ICF-specific: foam forms "shall be protected from sunlight and physical damage by the application of an approved exterior wall covering" (R608.4.3), with no fire minimum [EV-02]. The ICF evaluation reports say the same [M-21, EV-01].
The IBC's own exceptions describe SCIP's geometry, not ICF's: no thermal barrier is needed, and no NFPA 285 test, where foam "is covered on each face by not less than 1-inch (25 mm) thickness of masonry or concrete" (2603.4.1.1, 2603.5.5 Exception 2) [EV-03]. That is our reading of the text, not a ruling. ICF has concrete on one face only. For commercial buildings, ICF has an industry NFPA 285 basis (two passed tests, brick and ACM, extended by analysis to stucco, EIFS and fiber cement) [EV-07].
In the WUI the face matters. IWUIC Class 1 and 2 walls, California's WUI Code and Colorado's Class 2 walls must be noncombustible, ignition-resistant, fire-retardant-treated, 1-hour rated from the exterior side, or (in California) pass ASTM E2707 / SFM 12-7A-1 [EV-04, EV-05, N-29]. Colorado bars coverings that "melt, drip or delaminate" in the ignition-resistance test [N-29]. Cement stucco and fiber cement are noncombustible (ASTM E136); a thin polymer lamina is not [EV-12]. ICF's ratings are from the interior side unless at least 1 in. of concrete or masonry, or gypsum, covers the outside (ESR-2092 §4.4.2.1) [EV-01]. No ICF, EIFS or SCIP wall appears in the State Fire Marshal's listed-products handbooks (2022, 2024, 2025); unlisted assemblies may still comply prescriptively [EV-06].
No code sets a debris-impact or repair standard for wall finishes; the IRC's impact rule covers glazing only [EV-02].

What the tests and field evidence show.

Debris. Texas Tech fired 15-lb 2x4s at about 97–120 mph at four ICF walls. The concrete was never cracked or spalled, but vinyl, ¼-in. EIFS and the EPS behind them were holed through to the core; brick veneer cracked and the foam was only indented [EV-08, EV-09]. EIFS impact classes (ASTM E2486) top out above 150 in-lb, far below the roughly 4,000 in-lb of a 9-lb 2x4 at 34 mph (our arithmetic) [EV-10]. FEMA notes that lost wall coverings let wind-driven rain in, and that walls have no secondary water barrier comparable to a sealed roof deck [EV-36]. Its teams "frequently observed EIFS blow-off" in Charley, Ivan, Katrina, Ike and Michael, mostly on commercial EIFS over studs and gypsum; one Ivan case was EIFS dabbed onto cast concrete [FM-08, EV-23, EV-24, EV-25]. ICF foam is cast in and tied by the form webs, so that failure plane does not apply. Stucco had the lowest wall-covering damage in IBHS's 186-home Michael survey and "performed well overall" in Ian, where vinyl and fiber cement were the most often damaged at winds well below design [EV-30]; while vinyl and fiber cement failed widely in Michael (about 30% of vinyl on new homes) [EV-28, FM-11, FM-08, EV-29]. Debris still punctured stucco and lath (Charley, Ike, Harvey), and stucco on concrete cracked, spalled and admitted water in Maria [FM-02, EV-25, EV-26, EV-41]. The one documented ICF house in a major hurricane, Mexico Beach's "Sand Palace", lost "a few pieces of siding" (industry press) [EV-31].
Fire. In full-scale UL FSRI tests, EIFS (foam with a stucco skin) "can enable fire spread by separating and melting to burn in a pool at the base of the wall", while fiber cement withstood higher heat flux [EV-12]. Rendered EPS facades crack at about 30 kW/m² and the foam then burns inside the system (not a wildfire test) [EV-18]; a single protective layer let EPS reach softening temperature within 10 minutes in a furnace [EV-19]. Fox Blocks says its EPS "will melt when exposed to sustained temperatures over 300°F (149°C)" [EV-32]. Australian makers claim only BAL-29 for EPS ICF with acrylic render and steer flame-zone sites to other products (industry) [EV-20]. NIST: hardened cladding such as stucco or cement board is not challenged by low exposures, but a fully involved house nearby "will challenge almost all typical residential construction"; in its separation experiments a hardened target resisted ignition where an unhardened one ignited within 6 minutes [EV-15, EV-35]. IBHS's showcased Lahaina survivor had CMU walls [EV-17]. FEMA's Marshall Fire team found concrete block and brick "remained standing more often" and flagged exterior rigid foam as "typically more flammable" and in need of research [EV-13]; FEMA's draft wildfire builder's guide recommends CMU, cement board, masonry or three-coat stucco and lists EIFS among coverings to replace (draft, not yet released) [EV-14]. The only documented ICF foam melt in a wildfire is a trade-press case: an outbuilding in the 2015 Valley Fire where "the fire's intense heat melted the foam underneath the stucco"; the concrete core held and the owner refoamed and restuccoed it [EV-21]. ICF houses that survived the Camp, Witch Creek and Valley fires are reported only by industry, press or a student paper [EV-22]. Large survival studies log siding only as combustible or not, so no study separates ICF or foam-backed stucco [N-18, N-19, N-22].
Fairness to ICF. With a hard finish the picture improves a lot: brick kept the missile off the foam, fiber cement and cement stucco are noncombustible, and the concrete core stops burn-through whatever the finish [EV-08, EV-12]. The documented ICF failure mode is a damaged face with a local repair, not a lost house [EV-21].

The scores.

System (assumed face)	Fire on the face	Debris on the finish	Repair	Row score	Evidence badge
Grouted CMU, stucco direct on block or bare block	9	8	8	8.5	Extensive
SCIP, 1.25–2 in. reinforced shotcrete each face	8	8	8	8	Moderate
ICF, cement-based finish (three-coat stucco on lath, fiber cement or brick veneer): the scored assumption	7	6	6	6.5	Emerging
ICF, thin polymer-modified stucco or EIFS-type lamina on the EPS: variant	3	4	5	4	Moderate
FORTIFIED / ignition-resistant wood frame, fiber cement or stucco over wood sheathing	6	3	5	4.5	Extensive
SIPs, cladding over OSB skins	3	3	3	3	Emerging

CMU: noncombustible with no foam, the strongest Marshall Fire field record, grouted CMU intact after more than 30 hits at over 100 mph; stucco can crack or blow off [EV-13, EV-08, EV-23, EV-37].
SCIP: the code's "1 in. of concrete on each face" picture; 1-h ASTM E119 walls; passed the 9-lb missile (damage not described). No wildfire test, hail or field data beyond one builder's account of a SCIP house that stayed structurally sound in the 2015 Valley Fire [EV-39], and thin shotcrete could crack under long exposure (untested), so 8, not 9 [EV-03, EV-33, EV-08, EV-34].
ICF, hard finish: noncombustible covering and a concrete core, but foam behind it can still melt back under sustained heat and needs replacing; no exterior-side rating or WUI test [EV-01, EV-21].
ICF, thin finish: melts, delaminates and pool-burns; holed with the foam by a test missile; repair is a routine but specialist cut-out, foam plug, re-mesh and texture match, and melted foam can hide behind an intact-looking skin [EV-12, EV-08, EV-11, EV-40]. Fox Blocks calls thin-coat "typical" on ICF because it is cheaper; no ICF finish market data exists, and across all new US homes stucco, vinyl, fiber cement and brick each clad about a fifth to a quarter [EV-32, EV-38].
Wood frame: noncombustible cladding helps (Camp Fire: "stucco survived but framing burned out"), but the sheathing behind it burns at close range, and cladding over studs was perforated at 52–54 mph in the missile tests; FORTIFIED does not regulate siding [EV-16, EV-12, EV-08, N-18].
SIPs: the finish is the only layer over a combustible OSB skin and the foam core; FEMA notes debris damage "may require cutting out a section of the SIPs" [EV-27].

Thin evidence changes only the badge. The ICF hard-finish row is Emerging because no hard finish on ICF has a wildfire test or field survey; its score comes from the finish materials' own records.

Weights. 3% in every region plus 15% of the region's hurricane-wind and wildfire weight, rounded to the half percent (Section 2.4), computed on the base weights without termites. The other weights are scaled down in proportion.

Region	FL	GULF	TX	SATL	NEMA	NNE	APP	MIDS	LMW	UMW	NPL	SPL	ROCK	IMW	DSW	CA	PNW	AK	HI	PRVI
Exterior-face weight	12.5%	10%	8.5%	10.5%	9.5%	7.5%	4.5%	5.5%	5.5%	6%	7.5%	8%	13.5%	8%	12%	7.5%	4.5%	4%	12%	10.5%

Why ICF is scored with a cement-based finish, and what the thin finish would change. The IRC requires only "an approved exterior wall covering" over ICF, and thin polymer stucco is common [EV-02, EV-32]. We score each system as the region picks specify it, as with FORTIFIED wood frame and the engineered SCIP roof, so the ICF picks now name a cement-based finish. With the thin-coat variant, ICF's exterior face scores 4 instead of 6.5 and its totals fall by 0.1–0.3. Grouted CMU + concrete roof would then lead Florida, the Southern Plains, the Rocky Mountains and the Desert Southwest. The "How the scoring works" section lists the variant region by region.

Effect on the picks (cement-based ICF finish). Taken alone, this row changed four winners, all on the tie rule: SCIP took California, the Intermountain West and Puerto Rico / USVI, and ICF + wood-truss roof took the Pacific Northwest. The termite and latent-moisture changes made in the same revision (Section 4.8) then moved the Pacific Northwest and Hawaii to SCIP as well; Section 1.2 now follows the higher total. The 4 Nov 2026 correction retired the 0.2 tie window, so those three earthquake tie-breaks no longer overturn a higher total.

4. Evidence by hazard

4.1 Hurricane wind and windborne debris

Where the losses are. Hurricane wind is the #1 building-loss hazard in FL, LA and SC and #2 in NC, MS, AL, GA and TX. Florida is 30% of the US total [N-36].
Design speeds. ASCE 7-22 Risk Category II speeds on the Gulf and Florida coast run roughly 140–180 mph. Examples: Mobile, AL 146 mph; Destin, FL 152 mph; the Keys up to 180 mph [M-06, M-07].
How houses fail.
- Andrew: "concrete block and stucco (CBS) systems performed much better than all-wood-frame construction," but a "lack of vertical wall reinforcing" caused masonry failures [FM-01].
- Charley, Irma, Michael and Ian: post-FBC structures generally performed as expected. Damage concentrated in roof covers, openings and older homes [FM-02, FM-07, FM-08, FM-11]. In Ian, 90% of shingle roofs older than 7 years were damaged, against 28% of newer ones [FM-11].
Debris.
- Wood stud walls with plywood were perforated by the 15-lb 2x4 at 52–53 mph [CMU-04].
- Fully grouted reinforced CMU and ≥4-in. uniform ICF cores stop the 100-mph tornado missile [N-55, CMU-01].
- 8-in. CMU with only the rebar cells grouted was perforated at 104 mph [CMU-04].
- The finish is a separate question. The concrete in ICF walls stopped every 15-lb missile at about 100 mph, but vinyl, ¼-in. EIFS and the EPS behind them were holed through to the core; brick cracked and only indented the foam [EV-08]. FEMA's teams repeatedly find wall coverings stripped where the structure held, with water following, and stucco had the lowest damage frequency in Michael [FM-08, EV-28, EV-36]. Section 3.8 scores this for every system.
FORTIFIED.
- In Hurricane Sally (Alabama), FORTIFIED homes had loss frequency 55–74% lower, loss severity 14–40% lower and loss ratio 51–72% lower [FTD-09]. The IBHS press release rounds severity to 15–40% [M-10].
- New-build premium in Oklahoma: Roof 1.1%, Silver 1.6%, Gold 2.25% of sales price [COST-12].
- NIBS benefit-cost ratio for hurricane wind: 10:1 for adopting 2018 codes, 5:1 for exceeding them [COST-04].
- Louisiana's Regulation 136 requires FORTIFIED premium discounts on policies issued or renewed from January 1, 2027 [M-12].
Roofs. FEMA assessment teams found roof coverings the most commonly damaged element in Michael and Ian, and roof-to-wall connection failures in Irma (Section 3.7) [FM-07, FM-08, FM-11]. Reinforced concrete walls under a wood-truss roof therefore keep the roof as their weak point.
Implication. The wall choice matters most where design winds exceed about 140–150 mph, debris loads are high and code enforcement is weak. Otherwise the roof, openings and water intrusion dominate, which is why the scoring models the roof.

4.2 Storm surge and flooding

Inland flooding is the #1 building-loss hazard in 45 of 51 states plus DC [N-36].
V zones require buildings elevated on pilings or columns, with breakaway walls below the base flood elevation [N-37, M-01].
Elevation is what reduces loss. In Hurricane Ian, the average NFIP building claim was $46,033 for elevated homes against $134,398 for non-elevated ones [N-38].
Surge. In Katrina and Rita, "nearly all residential construction exposed to storm surge was completely destroyed" [FM-06]. One reinforced-concrete-frame-plus-ICF house survived amid dozens of destroyed homes [FM-05]. That is one documented case, not a statistic.
Materials.
- NFIP TB-2 accepts concrete, concrete block, closed-cell foam and solid structural wood studs below the flood elevation, but not OSB or paper-faced gypsum [FM-19].
- Concrete and EPS shells are therefore easier to dry out and reoccupy.
- Unfilled CMU cells can hold water [FM-19].
Helene (2024) caused at least 94 direct freshwater-flood deaths, 77 of them in North Carolina, and over 2,000 landslides [N-40]. In mountain flood country, siting matters more than structure.

4.3 Earthquake

Where. USGS says "nearly 75 percent of the U.S. could experience damaging earthquake shaking" [N-01]. The highest design motions are in Anchorage, Hilo, Los Angeles, San Francisco, Seattle and Salt Lake City [N-02]. California has 58% of US earthquake building loss [N-36].
Code paths by Seismic Design Category [N-04, N-05, M-21, M-22]:

System	Code path
Wood frame	Prescriptive through SDC D2; SDC E needs engineered design (AWC SDPWS)
IRC R608 concrete / ICF	Prescriptive in SDC A–C; beyond that, PCA 100 or ACI 318 with no SDC cap in the ICF evaluation reports
Masonry	R606.12 provisions for D0–D2
SIPs	SDC A–C only (IRC R610)
Cold-formed steel	D0–D2 must also comply with AISI S230
SCIP shear walls (ICC-ES)	Product-specific. ESR-5623 and ESR-3429: SDC A–B only. ESR-2435 (RSG 3-D/EVG): SDC C–F as ACI 318 Ch. 18 special structural walls, engineered [SCIP-27]
SCIP shear walls (DrJ TER, GCT)	SDC B–F claimed (R = 5), on confidential ASTM D7989 equivalency data; accepted case by case [SCIP-08]

Mass (code dead loads): wood wall 15 psf; cold-formed steel 14 psf; 8-in. masonry 80 psf; 6-in. concrete 85 psf [N-04]. GCT SCIP is about 35 psf [SCIP-08].
Field record.
- Houses "have traditionally performed fairly well in earthquakes because of their relative lightness" [N-06].
- Northridge: 24 of 25 building-caused deaths were in woodframe buildings, mostly soft-story apartments [N-07].
- Napa 2014: damage came from cripple-wall failures in older homes [N-09].
- Anchorage 2018: modern homes had cosmetic damage only, though shaking reached just 50–60% of design levels [N-08].
Guidance and retrofit. FEMA P-232 (2024) is the homebuilder's seismic guide [M-03]. California's Earthquake Brace + Bolt pays up to $3,000 for cripple-wall bracing [N-10].

4.4 Wildfire

Where (USFS risk to homes, state percentile): NV 1.00, CA 0.98, ID 0.96, WA 0.94, OR 0.92, UT 0.90, HI 0.88, AZ 0.86, OK 0.84, NM 0.82 [N-27].
What decides survival.
- Embers entering through eaves, vents, windows and the roof [N-18, M-13, N-103]. IBHS calls wind-blown embers the principal cause of building ignitions [M-13].
- Spacing: "each additional 10 ft of separation increased the likelihood of no damage by 7–13%," up to 30 ft [N-22].
- Zone 0 fuel: above 25% cover, damage likelihood was 87–100% [N-22].
- Building era: in the Camp Fire, homes built 1997 or later survived at 38.5%, against 11.5% for older homes. The before/after Chapter 7A difference (37% vs 44%) "was not statistically significant" [N-19].
The wall face. Noncombustible siding helps but is not decisive in the survival data, which never separates foam-backed stucco or ICF [N-18, N-22]. FEMA's Marshall team found block and brick walls "remained standing more often" and flagged exterior rigid foam for research; thin finishes over EPS melt back and pool-burn in full-scale tests [EV-13, EV-12]. Section 3.8 scores the face for every system.
Codes.
- California's 2025 WUI Code (effective Jan 1, 2026) [N-25]. Its walls must be noncombustible, ignition-resistant, 1-hour rated from the exterior side or tested to ASTM E2707 / SFM 12-7A-1 [EV-05].
- Colorado Wildfire Resiliency Code 2025 [N-29].
- Oregon R327 is available for local adoption from 2026 [N-28].
- Washington rescinded its 2021 WUI amendments in 2026 [N-30].
Cost. Building to wildfire code costs about 2–3% more [N-23].
Insurance. California's Safer from Wildfires programme requires discounts for a Class A roof, ember-resistant vents, enclosed eaves, multi-pane windows and Zone 0 [N-25].

4.5 Tornado

Where. In building loss, TX leads, then IL, MI, OH, IN, GA, IA and FL [N-36]. Tornado frequency is falling in TX, OK and NE Colorado and rising in MS, AL, AR, MO, IL, IN, TN and KY [N-53].
Intensity. 95% of 1950–2011 tornadoes were EF0–EF2 [N-55]. Designing for EF0–EF2 "could reduce tornado damage in as much as 85% of the impacted regions" [N-58].
Moore, OK (2014). A 135-mph design standard with garage-door, sheathing and connection upgrades costs about $1,500–2,500 per home. It "cannot be considered a storm shelter" [N-56].
Safe rooms.
- The test missile is a 15-lb 2x4 at 100 mph on vertical surfaces and 67 mph on horizontal ones [N-55].
- FEMA-funded residential safe rooms are designed for 250 mph [M-05].
- The current FEMA P-361 is the 5th edition (2024, based on ICC 500-2023) [M-04].
Codes after Joplin. NIST's Joplin investigation led to tornado-resistant design provisions in ASCE 7-22 [M-08, M-09].

4.6 Hail

Where. TX is #1 (30% of US hail loss), then CO, MO, WI, OK, NE, MN, OH [N-36]. "Nebraska, Colorado, and Wyoming usually have the most hailstorms" [N-59].
What it hits. "The majority of hailstone impacts were on the roofing system" [N-61]. Hail is a roof-covering problem first.
Remedy. IBHS shingle impact ratings and the FORTIFIED Hail Supplement [N-62, M-15, M-16]. Concrete-faced walls do not dent like vinyl or aluminium siding, but we found no study that measures this.

4.7 Snow, cold, freeze-thaw and energy

Snow. ASCE 7-22 adopted reliability-targeted snow maps. Loads rose in places such as Baltimore (25→60 psf), Chicago, Denver and Portland [N-63]. Collapse risk sits in roof framing, not walls [N-64].
Freeze-thaw.
- ACI 318 exposure classes F1–F3 require air entrainment [N-67]. Concrete must be detailed for freeze-thaw; being concrete alone does not solve it.
- Frost-protected shallow foundations are in the IRC [N-65].
Energy (wall R-values):

Wall	R-value (h·ft²·°F/Btu)	Source / caveat
ICF	24.1	Industry-funded test [ICF-11]
ICF (ORNL)	11.6 hot-box; mass-equivalent R-14.5 to R-20.5	[ICF-12]
3.5-in. SIP, whole wall	up to 14.1 (vs 9.6 for 2x4)	[SIP-15]
GCT SCIP (PSM60–PSM240; PSM80 = R-12)	9–34	Calculated (ASHRAE 90.1 method); no hot-box test [SCIP-08]
Emmedue (PSM60–PSM240; PSM80 = R-12.5)	9.7–34.7	Calculated; legacy report [SCIP-17]
EVG 3-D panel (40/100/40 mm)	about 5.7 to 12.5 (1.0–2.2 m²K/W)	BRANZ "approximate"; only the EPS was tested. Mild-steel truss wires give 1.0–1.7 m²K/W, stainless 1.7–2.2 [N-68]
8-in. CMU	1.7–2.6	[CMU-07]

The truss wires crossing the SCIP core are a thermal bridge that SCIP R-values must account for [N-68].
Is SCIP energy-efficient? Partly verified.
- A continuous EPS core with concrete on both faces is an insulated, airtight-capable shell, and the calculated R-values are competitive with framed walls.
- But no hot-box test of a whole SCIP panel (ASTM C1363 or ISO 8990) and no measured energy-use study of a SCIP house was found.
- BRANZ's figures for a comparable panel with mild-steel truss wires run roughly 40% below an EPS-only estimate. That is our arithmetic, and it suggests the calculated values in the US reports may be optimistic [N-68].
- Ask for a tested whole-wall R-value before relying on an energy figure.

4.8 Termites, insurance, humidity and latent moisture (revised 30 Oct 2026)

Termites: what the codes say.

Model code. IRC R318.4 (2009–2021), renumbered R305.4 in 2024: "In areas where the probability of termite infestation is 'very heavy' as indicated in Figure R305.4, extruded and expanded polystyrene, polyisocyanurate and other foam plastics shall not be installed on the exterior face or under interior or exterior foundation walls or slab foundations located below grade. The clearance between foam plastics installed above grade and exposed earth shall be not less than 6 inches (152 mm)." There are three exceptions:
1. "Buildings where the structural members of walls, floors, ceilings and roofs are entirely of noncombustible materials or pressure-preservative-treated wood."
2. An "approved method of protecting the foam plastic and structure from subterranean termite damage".
3. "the interior side of basement walls."

[TM-06, N-75] IBC 2603.8 is the same rule [N-74].

The map. It is Figure R301.2(7) in the 2018 IRC (R301.2(6) in 2009–2015, R318.4 in 2021, R305.4 in 2024) [N-74].
Our reading of exception 1. It covers the whole structure, roof included. An ICF or SCIP house with a concrete roof can qualify; ICF walls under wood trusses cannot. The code's concern is foam as a hidden path to wood.
Florida. FBC Building 2603.8 has the same text with no "very heavy" qualifier (its map is "Reserved"), so it applies statewide [TM-01]. The residential equivalent is FBC Residential R318.4. We confirmed that number from a 2023 Florida ICF evaluation, which says termite protection "may be required by the local jurisdiction in accordance with FBC 2603.8 or FBC-R R318.4"; the FBC-R wording itself was not retrieved [TM-05].
Florida inspection clearance. FBC Building 1403.8 requires 6 in. of clearance "between exterior wall coverings and final earth grade" "to provide for inspection for termite infestation" [TM-04].
Two corrections. FBC 1503.7 ("Protection against decay and termites") makes downspouts and condensate lines discharge at least 1 ft from the wall. FBC 1816 ("Termite protection") covers soil treatment, baits, wood treatment and certificates. Neither is a foam rule [TM-02, TM-03].

Termites: what the evidence shows.

Clemson. Termites "can tunnel through the insulation to reach the wooden portions of the structure, while hidden from view" [N-76]. In South Carolina:
- 43% of pest control firms "elected not to treat structures with below grade insulation";
- 12% had faced litigation over missed infestations;
- damage was found a mean of five years after construction;
- the pest-control association's 1993 advice was to "refuse the job" where the foam could not be removed.

These were mostly framed houses with foam sheathing.

UF/IFAS: termites "can easily tunnel through ... rigid foam board insulation and spray foam insulation" [TM-08].
Manufacturer (BuildBlock): termites "could burrow through the foam to reach and damage untreated wood", and "It is a misconception to say that ICF homes are termite proof" [TM-09].
ICF to the roof. The one ICF-to-roof case found is an InterNACHI account of termites that "had tunneled inside the EPS all the way to the roof". It gives no place, date or investigation, so it is an anecdote [TM-11].
Treated foam. Imidacloprid-treated EPS is code-recognised below grade in "very heavy" areas, but its label says it "is not sole protection" [N-74, TM-10]. PNNL recommends a 6-in. clearance and an inspection gap [N-78].
Federal mortgage practice. HUD accepts "steel, masonry or concrete" houses "with only minor interior wood trim and roof sheathing" with "no treatment needed" [TM-07].
No data for SCIP or SIPs. No SCIP termite test or field report was found, and no SIP field survey. SCIP's foam is encased in at least 1 in. of concrete on both faces [SCIP-04]; that is our reasoning, not a test.

Termites: how they are scored. The termite row is the mean of two part-scores, rounded to the half point:

System	Is the structure food?	Hidden path and inspectability	Termite row (old → new)	Badge
Grouted CMU + concrete roof	9 (only trim and partitions)	8 (tubes visible on block at grade)	9 → 8.5	Moderate
SCIP walls + roof	9	7 (foam encased on both faces; untested)	8.5 → 8	Emerging
ICF + concrete roof	9	4 (EPS hides galleries from the soil up)	8.5 → 6.5	Moderate
CMU + wood-truss roof	5 (the trusses)	8	9 → 6.5	Moderate
ICF + wood-truss roof	5 (the trusses at the end of the foam)	4	8.5 → 4.5	Moderate
Wood frame	2	5 (relies on the 6-in. inspection clearance)	3 → 3.5	Extensive
SIPs	2	3 (foam and OSB together)	3 → 2.5	Emerging

Cold-formed steel framing (not one of the seven mapped systems) is not food, and the IRC lists it as a termite control method [TM-06].

Termite weight by region. The weight is set by the region's predominant zone on the IRC map: 12% "very heavy", 7% "moderate to heavy", 3% "slight to moderate", 1% "none to slight". The previous edition's editorial termite weights (0–15%) are dropped.

Zone	Weight	Regions
Very heavy	12%	Florida, Gulf Coast, South Atlantic, California, Hawaii, Puerto Rico / USVI
Moderate to heavy	7%	Texas (very heavy on the coast), Northeast and Mid-Atlantic, Appalachia, Mid-South, Lower Midwest, Southern Plains, Desert Southwest
Slight to moderate	3%	Northern New England, Upper Midwest, Rocky Mountains, Intermountain West, Pacific Northwest
None to slight	1%	Northern Plains, Alaska

Basis for the zones.

HUD's Atlanta homeownership centre lists AL, FL, GA, MS, SC and Puerto Rico as "VERY HEAVY" and IL, IN, KY, NC and TN as "MODERATE to HEAVY" [TM-07].
Local Table R301.2(1) entries agree where we found them [TM-13]: Columbus, OH "Moderate to Heavy"; Spring Hill, TN "Heavy"; Goodyear, AZ "moderate to heavy"; Boise and Jefferson County, CO "slight to moderate"; Seattle "moderate"; Corona, CA "very heavy".
A pest-control map page puts Hawaii in Region I, "Very Heavy", and Alaska in Region IV, "None to slight" [TM-14].
The Northeast, Northern New England and Appalachia were read from the map only. The US Virgin Islands are assumed to match Puerto Rico. The map's own note says its "Lines defining areas are approximate only".

Effect. With termites weighted by zone and scored for hidden paths:

ICF + concrete roof loses 0.15–0.3 in the very heavy zones;
ICF under trusses drops further;
grouted CMU + concrete roof becomes the runner-up in 11 regions;
SCIP's encased foam helps it in California, Hawaii and Puerto Rico / USVI.

Insurance (checked 30 Oct 2026).

No decline or surcharge found. We found no US insurer underwriting guide, rating manual, SERFF filing or state regulator that declines or surcharges ICF homes, over termites, foam fire, moisture or anything else. We re-searched with different queries and source types (underwriting and surcharge terms, SERFF and rule manuals, Citizens construction definitions, homeowner accounts). The only US claim of declines is one anonymous insurance-forum post, which we do not credit. Anecdotes the other way say carriers rate ICF like concrete block.
Texas: an optional discount. The Texas Insurance Code lets insurers give an ICF discount: "an insurer may grant to an applicant a discount in the applicant's homeowners insurance premiums ... constructed with an insulating concrete form system" (§2006.002, from S.B. 581 of 2003) [IN-01].
UK: one lender refuses ICF. Abroad, one UK building society "will not lend on properties where the construction type is Insulating concrete form (ICF)" (November 2025) [IN-02]. It is a mortgage lender, not an insurer.
Where termites do touch insurance. Termite damage is generally treated as a pest-control matter, not an insured loss. Foam that hides termites can void pest-control contracts: Georgia's Department of Agriculture warns that undetected termites behind foam "could ultimately void pest control contracts", and UF/IFAS notes that many firms "will not place a warranty on homes with spray foam insulation" [TM-12, TM-08, N-76].
What this means. None of this enters a score. Ask the pest-control firm in writing whether the foam detail affects its warranty.

Humidity, decay and latent moisture.

Wood frame. Wood does not decay if kept below about 20% moisture content [MAT-07, N-80]. Behind stucco, "with OSB moisture becomes concentrated at OSB/building paper interfaces"; Lstiburek's remedy is a 3/8-in. ventilated gap [LM-05].
ICF. Building Science Corporation: "Because the concrete is installed between two vapor retarding layers, it will take several years for the concrete to dry to equilibrium", so no more than latex paint should go on the drywall [LM-01]. Modelling of ICF walls in cold, wet climates found that low-permeance layers on the outside hinder drying and drive moisture inward, and that the right insulation and vapor control avoids problems [LM-02]. A 2002 PCA study, as reported by ICF Builder, found after a year that "none of the building materials from any of the walls suffered from moisture damage" (industry-funded) [LM-03]. The wall holds nothing organic except the drywall paper.
Grouted CMU. "CMUs contain a large amount of water after construction ..., which can result in delayed and prolonged moisture damage of adjacent building materials, such as furring strips or fasteners", and an interior vapor retarder "can trap moisture and result in mold and organic decay" [LM-04].
SIPs. Roof-panel seams decayed in Juneau from leaking moist air [N-79].
SCIP. No hygrothermal or field moisture study was found. Each thin face dries to its own side, the truss wires can cool the inner face in cold climates [N-68], and the ~1-in. cover over galvanized mesh is thin [N-14, SCIP-11].
Not mold-proof. Concrete walls do not make a house mold-proof: drywall, roof decking and HVAC still need moisture control.

Moisture: how it is scored. The moisture row is the mean of three part-scores, rounded to the half point. Region weights are unchanged.

System	Latent moisture (trapped water, drying, condensation, mold)	Decay	Corrosion and freeze-thaw	Row (old → new)	Badge
ICF (both roofs)	7	9	8	8 → 8	Moderate
Grouted CMU (both roofs)	6	9	7	8 → 7.5	Moderate
SCIP	6	9	4	6 → 6.5	Emerging
Wood frame	4	3	5	4 → 4	Extensive
SIPs	2	3	5	3 → 3.5	Moderate

4.9 Expansive soils

See the USGS swelling-clay map [N-70].
In Texas and the Southwest, post-tensioned slabs "are the most common type of residential foundations," and fewer than 1% show unacceptable movement [N-72].
The wall material does not solve swelling soils; the geotechnical investigation, foundation type and drainage do. A rigid superstructure on a poorly designed foundation can crack badly [N-73].

5. System verdicts

5.1 One paragraph per system

ICF: the best national all-rounder.
- Strengths: reinforced concrete, continuous insulation, termite and decay resistance, flood tolerance, tornado-missile evidence, a large installer base and code paths to any SDC through ACI 318.
- Main limit: ICF solves the walls, not the roof. Scored both ways: with a concrete roof deck it leads 14 regions; with a wood-truss roof it scores up to 1.1 lower, leads nowhere and is the runner-up in 7.
- Termites: EPS is not food but hides galleries from the soil to the wood above, and a wood-truss roof is that wood (termite row 6.5 with a concrete roof, 4.5 under trusses). Specify termite-resistant EPS or a listed barrier at grade and a 6-in. inspection strip; in Florida, FBC 2603.8 requires protection below grade statewide [N-76, TM-01].
- Second limit: the EPS face needs a hard finish. Scored with three-coat stucco on lath, fiber cement or brick (6.5 on the exterior-face row); a thin polymer stucco scores 4 and would hand Florida, the Rocky Mountains and the Desert Southwest to grouted CMU (Section 3.8).
Reinforced CMU: scored the same two ways. Its concrete-roof version is the runner-up in 11 regions (Florida, the Gulf Coast, Texas, the South Atlantic, the Southern Plains, the Rockies, the Intermountain West, the Desert Southwest, California, Hawaii and Puerto Rico / USVI) and has the highest total in Hawaii and the Intermountain West, where SCIP takes the tie on earthquake. Stucco on grouted block gives it the best exterior face (8.5) and the best termite row (8.5), but fresh block stores water that harms furring and fasteners (moisture row 7.5) [LM-04].
Engineered wood frame (FORTIFIED, ignition-resistant): the earthquake specialist and the price and permitting leader.
- Strengths: low mass, ductile shear walls, the largest labour base and the most mature code paths. FORTIFIED has the best real-storm loss data and insurer recognition [FTD-09, M-10, M-12].
- Weaknesses: combustibility, termites, decay and weak ordinary roofs are real, and are addressed by the FORTIFIED and WUI packages.
SCIP: the most promising specialized shell.
- Strengths: wall-and-roof concrete continuity is unusual, the foam is encased in concrete on both faces (no exposed foam for fire, debris or termites), and GCT's HVHZ approval lists slab panels as well as walls (the wind and missile tests cover the wall panels). It leads California and Puerto Rico / USVI, both close calls.
- Limits: approvals are product-specific, and workmanship (shotcrete cover, wire laps, special inspection) matters unusually much [SCIP-04, SCIP-11]. Only one ICC-ES report (ESR-2435) reaches high-seismic SDCs [SCIP-27]. There is no 100-mph tornado-missile result for a single layer, no wildfire test, no termite test, no moisture study, no published roof uplift or roof-impact test, and no documented record in a real earthquake or hurricane.
- Peer-reviewed full-scale flexural tests of SCIP slabs exist [M-20].
Reinforced CMU: highly rational where block trades already exist (Florida, Hawaii, Puerto Rico). Resilience depends on full grouting, reinforcement and roof connections; "block house" does not automatically mean "hurricane-resistant house." FEMA's Michael team found unreinforced and stucco-surfaced CMU walls toppling in high wind [FM-08].
SIPs: the best cold-climate envelope. Weaknesses: moisture at joints and skins, combustible OSB skins, no debris resistance, and IRC limits of SDC C and 140–155 mph [SIP-02, N-79].
Cold-formed steel: a light, noncombustible, termite-proof frame and a reasonable runner-up in seismic and wildfire regions. Weaknesses: thermal bridging, coastal corrosion and under 0.5% of US new homes [N-82, N-83].
AAC: fits hot-dry and fire-prone regions, but US supply is uncertain and IRC SDC D limits apply [N-04, N-84–N-86].
MgO board is sheathing, not a structural system, and has a documented moisture/corrosion failure history [N-87–N-89].
Rammed earth is a niche, with a code path in New Mexico [N-17].
3-D printed concrete (ICON ESR-4652) is limited to SDC A–B [N-90].

5.2 Capability matrix

The numeric scores (0–10) and the Extensive, Moderate or Emerging data badge for every cell are in "How the scoring works" above the report. This table summarizes performance in dots: ●●● (strong) to ○ (weak or none). It is not an evidence grade; the evidence depth is the badge. ICF and CMU each have two versions, with a wood-truss roof and with a concrete roof deck.

Hazard	SCIP	ICF	Grouted reinforced CMU	SIPs	FORTIFIED wood frame	Cold-formed steel	AAC	MgO	Rammed earth
Hurricane wind pressure (walls)	●●● (±124 psf on GCT wall panels)	●●●	●●●	●●	●● (field data)	●●	●●	n/a (sheathing)	●
Windborne debris (9-lb)	●●● (80 ft/s on PSM 120 wall panels)	●●●	●●●	●● (≥6½ in.) [SIP-11]	● (fails unprotected)	● [MAT-01]	●● [MAT-01]	○	●●
Tornado missile (15-lb, 100 mph)	● (66-mph pass for a single panel; 100-mph pass for a double panel only)	●●●	●●● (fully grouted)	○	○ (needs safe room)	○	○	○	○
Roof wind performance (uplift, roof-to-wall)	●●● with its own roof slab (engineered; no uplift test published)	●● with wood trusses; ●●● with a concrete deck	●● with wood trusses; ●●● with a concrete deck	●●	●● (FORTIFIED roof data)	●●	●● (ESR-1652 roof panels)	○	○
Roof fire and ember path	●●● (noncombustible slab, engineered; no roof fire test; Class A justification needed)	● with wood trusses; ●●● with a concrete deck	● with wood trusses; ●●● with a concrete deck	●	●● (WUI package)	●●	●●●	●●	●●●
Earthquake (US SDC D–E)	●● (ESR-2435 to SDC F, engineered; full-scale shake table to 1.2 g; no field record)	●● (engineered); heavier with a concrete roof	●● (R606.12); heavier with a concrete roof	● (A–C only)	●●●	●●●	● [N-04]	n/a	○ [N-17]
Wildfire (wall)	●● (1–2 h rated; no wildfire test)	●●● (2–3 h rated in ESR-1269; EPS faces must be clad; no wildfire test) [M-21]	●●●	● (needs cladding)	●● (with WUI package)	●●	●●●	●●	●●●
Flood (damage-resistant)	●●	●●●	●● (fill cells)	○ (OSB)	● (studs OK, OSB not)	●●	●●	○ [N-87]	○
Termites (hidden paths to wood, whole house)	●●● (foam encased on both faces; untested)	●● with a concrete roof; ● with wood trusses (EPS hides galleries to the roof) [N-76]	●●● with a concrete roof; ●● with wood trusses	●	● (needs treatment and the 6-in. clearance)	●●●	●●●	●●	●●●
Latent moisture (trapped water, drying, condensation)	●● (thin faces; wire bridging; no study)	●● (concrete dries for years between foam layers; nothing organic) [LM-01]	●● (stores water; protect furring) [LM-04]	● (joint condensation) [N-79]	● (OSB behind reservoir claddings) [LM-05]	●●	●●	○ [N-87]	●●
Exterior face (fire, embers and debris on the finish; repair)	●●● (shotcrete over the foam on both faces; no wildfire test)	●● with a cement-based finish; ● with thin polymer stucco on the EPS [EV-08, EV-12]	●●● (stucco on block, no foam)	● (cladding on OSB)	● (cladding over combustible sheathing)	●●	●●●	n/a (sheathing)	●●●
Cold-climate energy	●● (calculated R; wire bridging; no hot-box test)	●●●	○ (unless insulated)	●●●	●●	● (thermal bridging)	●●	—	○
Labour / supply in US	○ (few plants)	●●	●●● (FL, Gulf, HI, PR)	●●	●●●	● (<0.5% of homes)	○	●	○
Relative cost (neutral data only)	High (1994: +63% walls) [COST-09]	+3–5% of price (2001) [COST-05]; a concrete roof deck adds more	Regional baseline in FL	+68% shell (industry, 2006) [COST-11]	+1–2.25% of price (Gold) [COST-12]	+14% (2002) [COST-10]	unknown	unknown	+20–30% (press) [N-94]

6. Uncertainties and where experts disagree

Do concrete walls reduce hurricane losses once code era is controlled?
- IBHS says the modern FBC gives near-FORTIFIED-Gold protection in much of Florida [FTD-21].
- MIT CSHub finds wood-frame damage about twice masonry's in Irma, but with small samples and no age control [MAT-02].
- Florida's wind-mitigation credits list roof, opening and connection features, but not wall material [COST-21].
- No insurer study isolates wall type at equal code era.
Wall or roof? FEMA assessment teams put most losses at roof coverings, decks, openings and water entry. Concrete advocates answer that a concrete roof eliminates those. Both are right: the roof is where a monolithic shell earns its keep, and most ICF and CMU houses still have wood roofs. This edition scores both versions, and the 40% roof share inside the wind family, and the 40% of fire weight moved to the ember row, are editorial judgements that move close calls.
SCIP field record.
- There is no independent post-disaster engineering report on SCIP houses.
- Andrew-era survival stories (Liberty City Habitat houses, one Homestead house) rest on a 1992 newspaper article repeated in vendor-hosted material [M-27].
- Claims about Hurricane Hugo with a quoted 195-mph wind are not credible as stated and are not used.
- Earthquakes: no EERI, EEFIT or Build Change report on Ecuador 2016 mentions SCIP houses, and no engineering account was found for Chile 2010, Mexico 2017, Turkey 2023 or the Italian, Central American, Haitian, Nepalese or Albanian earthquakes [SCIP-36, SCIP-37, SCIP-38]. The founder's Ecuador observation (Section 3.3) is the only first-hand account we have, and it is not a survey.
SCIP seismic path.
- ICC-ES ESR-2435 sends SDC C–F walls to ACI 318 special-structural-wall design [SCIP-27]. Whether a thin shotcrete-and-mesh wythe can meet every Chapter 18 detailing rule without much added reinforcement is an engineering question the report does not answer.
- A FEMA P-695 study supports R = 3.5 only for one- and two-storey 3-D panel buildings [SCIP-29]. ESR-2435 and GCT's TER instead point SDC D–F design to the special-wall coefficients (R = 5) [SCIP-27, SCIP-08]. Engineers may reasonably prefer the more conservative value.
- GCT's DrJ TER claims SDC B–F through ASTM D7989 equivalency on confidential data [SCIP-08]. Engineers will reasonably disagree on whether that is enough in high-seismic jurisdictions without peer review.
SCIP durability.
- Cover over galvanized mesh is 1 in. minimum (ESR) or ¾ + ¼ in. (GCT install guide), while ACI's weather-exposure cover for small bars is 1½ in.
- Long-term corrosion in salt spray is outside the scope of every SCIP evaluation report [N-14, SCIP-11].
SCIP thermal performance. GCT's and Emmedue's R-values are calculated. BRANZ's approximate values for a comparable 40/100/40 panel run from about R-5.7 to R-12.5, depending on truss-wire material and density [N-68]. No whole-panel hot-box test was found.
Cost. No neutral, current, apples-to-apples life-cycle cost database exists for SCIP, ICF, CMU, SIPs or steel. Treat any "X% more than wood" claim cautiously. BuiltStrong should collect real permitted-project bids by ZIP code.
Approval status changes.
- ESR-3429 (Schnell) is due for renewal in November 2026; ESR-2435 (RSG) in April 2027; RSG's Miami-Dade NOA expires March 2030 [N-14, SCIP-27, SCIP-28].
- FSS's FL46983 revision (R1) is in "Re-Apply" status.
- Florida's 9th-edition FBC takes effect December 31, 2026, which triggers product re-approval [N-43].
- Recheck SCIP approvals before each page update.
Seismic: light vs strong. Light frame's low mass versus concrete's stiffness is a real debate. Chilean concrete-wall buildings performed well in 2010 [N-15], and the Eucentre SCIP test exceeded its design collapse level [N-11]. SCIP sits between the two: a concrete shell at under half the weight of solid concrete walls. US codes and field experience still favour wood for houses [N-04, N-06].
Wildfire.
- Some studies weight structure features above defensible space [N-18]; others find defensible space "as important as building construction" [N-100].
- Chapter 7A's effect is not proven in the Camp Fire data [N-19].
- A concrete structure raises the baseline, but a poorly detailed concrete house can still burn.
- Exposed foam. FEMA's Marshall team flags exterior rigid foam as needing research, and no study separates foam-backed stucco or ICF in wildfire survival data [EV-13, N-22]. The exterior-face scores for ICF rest on lab tests of the finishes (much of it on frame walls), a few anecdotes and the codes. A wildfire test (ASTM E2707) of ICF with three-coat stucco, and of SCIP, would settle a row that decides Florida, the Southern Plains, the Rocky Mountains and the Desert Southwest under the thin-finish variant.
Termites, foam and insurance.
- The hidden-path evidence is strong for foam generally: Clemson's surveys, UF/IFAS and the codes [N-76, TM-08, TM-06]. For ICF specifically it rests on manufacturer statements, trade press and one inspector's anecdote of termites reaching the roof inside the EPS [TM-09, TM-10, TM-11].
- No SCIP termite test exists; its score of 8 is reasoning from encased foam.
- The zone weights (12/7/3/1%) are editorial.
- No US insurer was found to decline or surcharge ICF homes. If one does, that belongs in the price and permitting layer, not the structure score.
Latent moisture. ICF moisture evidence is one industry-funded year-long test and modelling [LM-02, LM-03]. There is no independent field survey of mold in ICF homes, and no moisture study of SCIP at all. The part-scores are engineering judgement from the mechanisms BSC and the Building America advisor describe [LM-01, LM-04, LM-05].
Tornado philosophy. One camp designs homes for EF2 and shelters occupants from the rest [N-58]. The other hardens the whole house, which only a concrete shell with a concrete roof can do. FEMA backs the first approach plus safe rooms [N-55].
Codes are minimums. FEMA's Building Codes Save study estimates that modern codes avoid at least $32 billion in losses over about 20 years, growing to more than $132 billion by 2040 [M-02]. But codes target life safety, not rapid reoccupancy, and adoption and enforcement vary widely [N-41, M-11].
NRI model change. Inland-flood loss jumped with v1.20's probabilistic model. Because the NRI's "hurricane" is wind only, other experts would weight surge and hurricane higher for coastal counties [N-34, N-35].
Insurance signals change faster than structural science. Discounts and grants (FORTIFIED mandates in AL and LA, the CA Safer from Wildfires programme) should be stored as a separate, regularly refreshed data layer [N-25, N-44, M-12].
Data we could not close:
- the full text of FEMA P-908 (Joplin) (P-2020, Maria, was read for the exterior-face revision [EV-41]);
- a SCIP roof uplift test, a SCIP roof debris-impact test (the roof is scored on an engineering basis meanwhile), a Florida ruling on whether a foam-core panel's 2-in. face meets FBC §1626.4, and a study isolating roof-to-wall connections from the other FORTIFIED requirements;
- official 2024 code text;
- AISI S230 limits;
- Florida-specific framing shares;
- Louisiana and Mississippi FORTIFIED statute texts;
- the underlying lab reports behind DrJ TERs, which are not public;
- the exact FBC Residential 2023 R318.4 wording, Texas's ICF discount rule, the ISO homeowners insect-exclusion text and the 2002 PCA ICF moisture report.

7. Sources

All URLs were accessed 2026-10-28.

7.1 SCIP approvals, evaluations and manufacturer tests

SCIP-01 Florida Product Approval FL46983-R0 (Fortified Structural Solutions LLC; "Concrewall USA SCIP"; HVHZ: No). https://www.floridabuilding.org/pr/pr_app_dtl.aspx?param=wGEVXQwtDquRQeeNyKB1%2fPLm%2bA22B5PPtO428M0XJZDJefix9Ic6iw%3d%3d
SCIP-03 ICC-ES ESR-5623, earlier edition under the Concrewall USA name (Florida copy). https://www.floridabuilding.org/upload/PR_Tech_Docs/FL46983_R0_AE_ESR-5623.pdf
SCIP-04 ICC-ES ESR-5623, FSS SCIP (Jan 2026, rev. Feb 2026). https://cdn-v2.icc-es.org/wp-content/uploads/report-directory/ESR-5623.pdf
SCIP-06 Florida Product Approval FL30050-R4, Gulf Concrete Technology (HVHZ: Yes; ±124 psf). https://www.floridabuilding.org/pr/pr_app_dtl.aspx?param=wGEVXQwtDqtSgVKbL0u0YeDREKSc6o7%2bF13GQ5QOtnTc1iXWsgyuPA%3d%3d
SCIP-08 DrJ TER 1202-12, GCT Insulated Concrete Panels (rev. Mar 2026). https://www.drjcertification.org/report/download/69 ; Florida copy: https://www.floridabuilding.org/upload/PR_Tech_Docs/FL30050_R4_AE_260305_TER1202-12_FPA%20Sealed.pdf
SCIP-11 GCT installation instructions (FL30050-R4). https://www.floridabuilding.org/upload/PR_Instl_Docs/FL30050_R4_II_260305_FL30050_TER1202-12_Install_Instrucs.pdf
SCIP-12 DrJ TER 1201-04, GCT fire resistance. https://www.drjcertification.org/report/download/68
SCIP-14 USAspending recipient record, Gulf Concrete Technology LLC. https://api.usaspending.gov/api/v2/recipient/2c909920-a804-267c-6590-c54f5f78afd5-P/
SCIP-17 DrJ TER 2110-01, Emmedue (past renewal, "LEGACY"). https://www.drjcertification.org/report/download/2011
SCIP-18 ETA 20/0976, Emmedue Building System (IETcc). https://dit.ietcc.csic.es/wp-content/uploads/2021/06/ETA-20-0976_2021-06-01_EMMEDUE-BUILDING-SYSTEM_signed.pdf
SCIP-24 Miami-Dade Product Control, NOA applicant search. https://www.miamidade.gov/building/pc-search_app.asp
SCIP-25 / CMU-04 / ICF-07 Texas Tech / NIST debris-impact summary (2006). https://www.depts.ttu.edu/nwi/research/DebrisImpact/Reports/TTU_Final_NIST_Report_numbered.pdf
SCIP-26 Texas Tech WISE, Emmedue M2 projectile test (2005; manufacturer-commissioned, third-party copy). https://www.oceancityinvestments.com/Test_-_Wind.pdf
N-14 ICC-ES ESR-3429, Schnell Home Sandwich Panels (rev. May 2026). https://icc-es.org/wp-content/uploads/report-directory/ESR-3429.pdf
N-11 Ricci et al., shaking-table test of a 3-storey sandwich-panel building, 15 WCEE (2012), paper 0884. https://www.iitk.ac.in/nicee/wcee/article/WCEE2012_0884.pdf
N-12 Bakhshi, Kabir & Rezaifar, Scientia Iranica 16(3), 2009. https://scientiairanica.sharif.edu/article_3108.html
N-13 ETA-23/0089, Green Walls SpA (IETcc, 2023). https://dit.ietcc.csic.es/wp-content/uploads/2024/01/report_ETA-23-0089_v1_2023-12-19_GREEN-WALLS.pdf
N-68 BRANZ Appraisal No. 750 (2023), EVG 3D Building System. https://www.branz.co.nz/assets/certified-products/750_2023.pdf
M-20 Acharya, Gurung & Mashal, "Full-scale flexural testing of slabs made of modular structural concrete insulated panels," PCI Journal 67(2), 2022, doi:10.15554/pcij67.2-03. https://www.pci.org/PCI/Publications/PCI_Journal/Issues/2022/March-April/Full-scale-flexural-testing-of-slabs-made-of-modular-structural-concrete-insulated-panels.aspx
M-27 "Case Study in Sustainability: Crystal Beach SCIP House" (FEMA-styled draft, 2009, with editorial notes; not in the published FEMA P-757 Hurricane Ike report; vendor-hosted; anecdotal Andrew/Hugo claims). https://www.tiahco.com/wp-content/uploads/2021/03/FEMAArticle-CrystalBeachHouse-061209.pdf
M-28 Concrete Products, Emmedue panels and ICC-ES ESR-2037 (Jul 19, 2010). https://concreteproducts.com/index.php/2010/07/19/blank-10656869/
SCIP-27 ICC-ES ESR-2435, RSG 3-D Structural Panel (3-D/EVG Panels), Re-Structure Group LLC (reissued Apr 2026, rev. Oct 2026; SDC C–F as special structural walls; fire Table 1; LA, CA and FL supplements). https://icc-es.org/wp-content/uploads/report-directory/ESR-2435.pdf
SCIP-28 Miami-Dade NOA 25-0421.03, RSG 3-D Structural Panel System (approved May 8, 2025; expires Mar 5, 2030). https://www.miamidade.gov/building/library/productcontrol/noa/25042103.pdf
SCIP-29 Mashal & Filiatrault, "Quantification of seismic performance factors for buildings incorporating 3D construction system," 15 WCEE (2012), paper 1693. https://www.iitk.ac.in/nicee/wcee/article/WCEE2012_1693.pdf
SCIP-30 Gurung, Experimental Investigations of Full-Scale MetRock SCIPs, MS thesis, Idaho State University (2019). https://etd.iri.isu.edu/GetFile.aspx?exid=1797&etid=1&mtid=1&inpage=true
SCIP-31 Universidad Católica de Santa María, thesis comparing confined masonry and a non-conventional panel system (2024), pp. 48–49 (Peruvian approvals RM 045-2010-VIVIENDA and RM 027-2013-VIVIENDA; secondary). https://repositorio.ucsm.edu.pe/bitstreams/44dd53bd-4e0e-4ab2-8eec-6e44db9eca87/download
SCIP-32 Palermo et al., "Seismic response of a full-scale sandwich-panel building," Engineering Structures 76 (2014): 75–89, doi:10.1016/j.engstruct.2014.06.024 (abstract read). https://doi.org/10.1016/j.engstruct.2014.06.024
SCIP-33 GCT, "Structural Insulated Panel Components – SCIPs" pamphlet (manufacturer, Jul 2023). Public link unavailable when checked on November 8, 2026.
SCIP-34 Carmelo, structural concrete insulated panels page (GCT parent; manufacturer). https://www.carmelo.com/en/carmelo-structural-concrete-insulated-panels/
SCIP-35 Emmedue, "Saudi Arabia fire tests" (manufacturer, Dec 2022). https://www.mdue.it/en/saudi-arabia-fire-tests.php
SCIP-36 EERI, Earthquake Reconnaissance Team Report: M7.8 Muisne, Ecuador (2016). https://discovery.ucl.ac.uk/id/eprint/1524378/1/EERI-Ecuador-2016.pdf
SCIP-37 EEFIT, The Muisne, Ecuador Earthquake of 16 April 2016 (IStructE, 2018). https://www.istructe.org/IStructE/media/Public/Resources/report-eefit-mission-muisne-ecuador-20180910.pdf
SCIP-38 Build Change, Ecuador post-earthquake reconnaissance report (Aug 2016). https://get.buildchange.org/wp-content/uploads/2016/08/2016-08-31_Ecuador-Post-EQ-Recon-Rpt_LowRes.pdf
SCIP-39 Pavese (Univ. of Pavia / TREES Lab), "Comportamento statico di solai e pareti EMMEDUE", in Prove sismiche su tavola vibrante del sistema EMMEDUE, ENEA C.R. Casaccia, 27 Nov 2008 (slides; static and cyclic tests only). https://www.yumpu.com/it/document/view/15186791/comportamento-statico-di-solai-e-pareti-emmedue-enea
SCIP-40 Secondary summary of PUCP Laboratorio de Estructuras, "Evaluación experimental del Sistema constructivo 'M2'" (A. San Bartolomé), report INF-LE-350-08, March 2009, slides 58–59 of "Paneles estructurales de EPS". The original report was not found online. https://www.slideshare.net/slideshow/paneles-estructurales-de-eps/234707920
SCIP-41 Ricci, Sistemi strutturali cellulari a pareti portanti in cemento armato gettato in opera realizzate con la tecnologia del pannello di supporto in polistirene, PhD thesis, University of Bologna (2012), §1 (names the Nidyon NYSP panels and Nidyon Costruzioni as funder of the Eucentre test). https://amsdottorato.unibo.it/id/eprint/4886/1/Ricci_Ilaria_Tesi.pdf
SCIP-42 Bassotti, Speranzini & Vignoli, "Experimental results and numerical simulation of a concrete sandwich panel system proposed for earthquake resistent buildings," ISEC-02, Rome, 23–26 Sep 2003, pp. 1165–1175 (abstract read; no full text online). https://research.unipg.it/handle/11391/140655
SCIP-44 Fortified Structural Solutions (formerly Concrewall USA), demo-day presentation ("Fortified Slide Show", Canva deck, June 2026). Manufacturer presentation, not a lab report: no test report, specimen drawing or missile specification is attached. Used only for the claims checked against SCIP-01 and SCIP-04.
SCIP-43 BMTPC (India), Performance Appraisal Certificate PAC 1010-S/2014, EMMEDUE Advanced Building System (8 Jan 2014), §I-3 and §III-2. https://bmtpc.org/DataFiles/CMS/file/PDF_Files/26_PAC-EMMEDUE.pdf

7.2 ICF, CMU, SIPs and other systems

M-21 ICC-ES ESR-1269, Amvic ICF. https://icc-es.org/wp-content/uploads/report-directory/ESR-1269.pdf
M-22 ICC-ES ESR-1147, Integraspec ICF. https://icc-es.org/wp-content/uploads/report-directory/ESR-1147.pdf
ICF-05 HUD/PCA prescriptive method for ICFs. https://www.huduser.gov/publications/pdf/icf_2ed.pdf
ICF-11 ICFMA thermal study (industry). https://icf-ma.org/download/icfma-thermal-study/?wpdmdl=254
ICF-12 ORNL whole-wall thermal calculator. https://web.ornl.gov/sci/buildings/docs/Whole-Wall-Thermal-Performance-Calculator.pdf
N-97 Fox Blocks, ICC 500 missile test announcement (manufacturer, 2022). https://www.foxblocks.com/blog/icc-500-missile-impact-testing
CMU-01 FEMA P-361 (2021), Texas state mirror. https://newtools.cira.state.tx.us/upload/page/12213/docs/SR/FEMA%20Publication%20P-361.pdf
CMU-07 CMHA TEK 06-02C. https://www.cmha.org/?create-tek-pdf=1&post_id=521983&filename=TEK%2006-02C
SIP-02 2025 California Residential Code §R610 (SIPs). https://gocodebook.com/library/us/ca/ca-residential-code/chapter-6-wall-construction/section-r610-structural-insulated-panel-wall-construction
SIP-11 FL28131-R3 SIP wall evaluation. https://www.floridabuilding.org/upload/PR_Tech_Docs/FL28131_R3_AE_SIPA083118-20r3%20SIP%20Wall%20Evaluation%20Report.pdf
SIP-15 ORNL whole-wall rating for SIPs. https://www.sips.org/documents/ORNL-Whole-Wall-Rating-for-SIPs-Steady-State-Thermal-Analysis.pdf
SIP-16 ORNL SIP vs wood-frame blower-door tests. https://www.sips.org/documents/blowerdoorheat.pdf
N-79 SIPA report on the Juneau roof issue (industry, 2002). https://www.sips.org/blog/sipa-report-on-the-juneau-alaska-roof-issue
N-81 2025 CRC §R603 cold-formed steel. https://gocodebook.com/library/us/ca/ca-residential-code/chapter-6-wall-construction/section-r603-cold-formed-steel-wall-framing
N-82 NFIP Technical Bulletin 8, corrosion protection. https://www.fema.gov/sites/default/files/2020-07/fema_tb8_corrosion_protection_metal_connectors_coastal_areas.pdf
N-83 NAHB Eye on Housing, framing shares 2024. https://eyeonhousing.org/2025/08/wood-framed-home-share-increased-in-2024/
N-84 ICC-ES ESR-1652, Aercon AAC panels. https://icc-es.org/wp-content/uploads/report-directory/ESR-1652.pdf
N-85 Concrete Products, Aercon/Litecrete AAC plant plans (2024). https://concreteproducts.com/index.php/2024/02/08/aercon-litecrete-j-v-outlines-autoclaved-aerated-concrete-panel-plant-ambitions/
N-86 Kerr County Lead, Megaacrete plant cancelled (2025). https://kerrcountylead.com/in-a-major-economic-blow-megaacrete-cancels-kerrville-manufacturing-plant-deal/
N-87 Rode (DTU), moisture challenges including MgO boards (2021). https://research.tuni.fi/app/uploads/sites/115/2026/02/rode-some-recent-challenges-and-peculiarities-in-moisture-practices.pdf
N-88 CCMC/NRC MgO board advisory update (2018). https://sboa.sk.ca/ccmc-mgo-board-advisory-update/
N-89 ICC-ES, codes and standards for MgO products (2024). https://icc-es.org/wp-content/uploads/ES-MgOBPA-Presentation-July-2024.pdf
N-90 ICC-ES ESR-4652, ICON 3D-printed walls. https://icc-es.org/wp-content/uploads/report-directory/ESR-4652.pdf
N-17 14.7.4 NMAC, New Mexico Earthen Building Materials Code. https://www.srca.nm.gov/parts/title14/14.007.0004.pdf
N-94 San Antonio Report, rammed-earth cost (press, 2026). https://sanantonioreport.org/what-is-rammed-earth-and-could-it-make-affordable-housing-more-affordable-to-own/

7.3 Codes and standards (ICC, ASCE, Florida)

N-04 2018 Seattle Residential Code (IRC-based), Ch. 3 and Ch. 6. https://www.seattle.gov/documents/Departments/SDCI/Codes/SeattleResidentialCode/2018SRCChapter3.pdf ; https://www.seattle.gov/documents/Departments/SDCI/Codes/SeattleResidentialCode/2018SRCChapter6.pdf
N-05 2021 IRC R608 (UpCodes). https://up.codes/s/exterior-concrete-wall-construction
M-23 ICC 2024 International Residential Code. https://codes.iccsafe.org/content/IRC2024V1.0
N-03 FEMA, 2024 Seismic Design Category maps; ASCE 7 Hazard Tool. https://www.fema.gov/sites/default/files/documents/fema-seismicdesigncategorymaps-july2024.pdf ; https://asce7hazardtool.online/
M-24 ASCE Hazard Tool overview. https://www.asce.org/publications-and-news/asce-hazard-tool
M-06 FEMA, ASCE 7-22 wind highlights fact sheet (2022). https://www.fema.gov/sites/default/files/documents/fema_asce-7-22-wind-highlights_fact-sheet_2022.pdf
M-07 NIST-WindMaps ASCE 7-22 Risk Category II wind-zone layer. https://services1.arcgis.com/ZWOoUZbtaYePLlPw/ArcGIS/rest/services/windzones_cat2_asce7_22_jun21/FeatureServer
N-63 Maguire, "Ground Snow Loads for ASCE 7-22," STRUCTURE, 2022. https://www.structuremag.org/article/ground-snow-loads-for-asce-7-22/
N-67 NRMCA guide to ACI 318 exposure classes. https://www.nrmca.org/wp-content/uploads/NRMCA_GuideToSelectingExposureClasses.pdf
N-74 ICC-ES ESR-2918 (termite map and IRC section mapping). https://www.icc-es.org/wp-content/uploads/report-directory/ESR-2918.pdf
N-75 S.C. Code Regs. 8-1216 (IRC R318.4 adoption). https://www.law.cornell.edu/regulations/south-carolina/R-8-1216
N-96 Florida Building Commission DS 2012-81 (FBC §1626.4). https://www.floridabuilding.org/fbc/commission/FBC_0113/Structural_TAC/DS2012-81_analysis_R2.htm
LF-13 FBC staff analysis DS 2017-030 (§1626 missile). https://www.floridabuilding.org/fbc/commission/FBC_0817/DEC_Statements/Staff_Analysis_DS_2017-030.htm
N-43 Florida Building Commission, FBC 9th edition product-approval changes. https://www.floridabuilding.org/fbc/Committees/product_approval/2026_Changes/2026_PA_Presentation_Chat.pdf
OPN-02 Florida OIR-B1-1802 wind mitigation form. https://floir.gov/docs-sf/property-casualty-libraries/product-review/all-forms/form-oir-1802-adopted-version-(no-watermark).pdf?Status=Master
N-25 California Dept. of Insurance, Safer from Wildfires; 2025 California WUI Code overview. https://www.insurance.ca.gov/01-consumers/200-wrr/Safer-from-Wildfires.cfm ; https://www.sandiego.gov/sites/default/files/2025-12/item-3-2025-cwui-code-overview-ppt_0.pdf
N-28 Oregon BCD, R327 amendment. https://www.oregon.gov/bcd/codes-stand/Documents/23orsc-r327-amend.pdf
N-29 2025 Colorado Wildfire Resiliency Code. https://tlcco.specialdistrict.org/files/df0134012/2025+Colorado+Wildfire+Resiliency+Code.pdf
N-30 Washington WSR 26-07-075. https://lawfilesext.leg.wa.gov/law/wsr/2026/07/26-07-075.htm
N-101 Shums Coda Associates for Montana DLI, 2024 IBC significant changes (IBC 104.2.3–104.2.3.7: alternative materials, research reports, peer review). https://bsd.dli.mt.gov/_docs/Training/2024-IBC-Significant-Changes.pdf
N-102 ICC, 2024 IRC significant changes (R104.11 renumbered R104.2.2). https://www.iccsafe.org/wp-content/uploads/Session-1-and-31-2024-IRC-Significant-Changes.pdf

N-104 IBC 1505.2 Class A roof assemblies (2021 model text, as printed in the Seattle Building Code, Ch. 15; exposed concrete roof deck listed as Class A). https://www.seattle.gov/documents/Departments/SDCI/Codes/SeattleBuildingCode/2021SBCChapter15.pdf
N-106 Florida Building Code, 8th Edition (2023), Building, Ch. 16 (§1609.1.2, §1616.1.1, §1620, §1625.3, §1626, §1626.4), Ch. 15 (§1514, §1515, §1516.2.1), Ch. 26 (§2603.4.1.1), §104.11; read on UpCodes (third-party copy; codes.iccsafe.org was not readable). The 9th Edition (2026) takes effect 31 Dec 2026, and the Commission's Analysis of Changes lists no change to these sections [N-107]. https://up.codes/viewer/florida/fl-building-code-2023/chapter/16/structural-design
N-107 Miami-Dade RER, FBC 9th (2026) Edition effective date notice; FBC Analysis of Changes, 9th Edition (2026) Building. https://www.miamidade.gov/resources/economy/building/documents/fbc-ninth-code-change.pdf ; https://www.floridabuilding.org/fbc/thecode/2026_Code_Development/Analysis_Code_Changes/Analysis_of_changes_9th_Edition-2026-FBC-Building.pdf
N-105 IRC R902.1 roof covering materials (2021 model text, as printed in the Seattle Residential Code, Ch. 9). https://seattle.gov/documents/Departments/SDCI/Codes/SeattleResidentialCode/2021SRCChapter9.pdf

7.4 FEMA

N-33 / M-25 FEMA National Risk Index (v1.20 now hosted in RAPT). https://www.fema.gov/flood-maps/products-tools/national-risk-index ; https://www.fema.gov/emergency-managers/practitioners/resilience-analysis-and-planning-tool
N-34 NRI Technical Documentation v1.20. https://www.fema.gov/sites/default/files/documents/fema_national-risk-index_technical-documentation.pdf
N-35 NRI Data Version and Update Documentation. https://www.fema.gov/sites/default/files/documents/fema_national-risk-index_data-version-update-documentation.pdf
N-36 NRI States Expected Annual Loss feature service (December 2025). https://services.arcgis.com/XG15cJAlne2vxtgt/arcgis/rest/services/National_Risk_Index_States_Expected_Annual_Loss/FeatureServer/0
M-02 FEMA Building Codes Save study. https://www.fema.gov/emergency-managers/risk-management/building-science/building-codes-save-study
N-41 FEMA Building Code Adoption Tracking (BCAT). https://www.fema.gov/emergency-managers/risk-management/building-science/bcat
M-03 FEMA P-232, Homebuilders' Guide to Earthquake-Resistant Design and Construction (Sept 2024). https://www.fema.gov/sites/default/files/documents/fema_p-232_september2024.pdf
N-55 FEMA P-361, Safe Rooms for Tornadoes and Hurricanes, 4th ed. (2021). https://www.fema.gov/sites/default/files/documents/fema_safe-rooms-for-tornadoes-and-hurricanes_p-361.pdf
M-04 FEMA P-361, 5th ed. (2024). https://www.fema.gov/sites/default/files/documents/fema_rsl_fema-p-361-safe-rooms-for-tornadoes-and-hurricanes_042025.pdf
M-05 FEMA P-320, Taking Shelter from the Storm (Dec 2024). https://www.fema.gov/sites/default/files/documents/fema_rsl_fema-p-320-taking-shelter-from-the-storm_042025.pdf
M-01 FEMA P-499, Home Builder's Guide to Coastal Construction. https://www.fema.gov/sites/default/files/2020-08/fema499_2010_edition.pdf
N-39 FEMA P-55, Coastal Construction Manual, Vol. I. https://www.fema.gov/sites/default/files/2020-08/fema55_voli_combined.pdf
M-19 FEMA NFIP Technical Bulletins index. https://www.fema.gov/emergency-managers/risk-management/building-science/national-flood-insurance-technical-bulletins
FM-19 NFIP TB-2, flood damage-resistant materials (2025). https://www.fema.gov/sites/default/files/documents/fema_tb_2_flood_damage-resistant_materials_requirements_01-22-2025.pdf
N-37 NFIP Technical Bulletin 5, free-of-obstruction. https://www.fema.gov/sites/default/files/documents/fema_free-of-obstruction-technical-bulletin_2021052721.pdf
N-38 FEMA, NFIP claims analysis, Hurricane Ian. https://www.fema.gov/sites/default/files/documents/fema_rm-nfip-claims-analysis-ian-case-study-12-2023.pdf
FM-01 FIA-22, Building Performance: Hurricane Andrew. https://www.preventionweb.net/files/1548_BuildingPerformance.pdf
FM-02 FEMA 488, Hurricane Charley MAT. https://www.fema.gov/sites/default/files/2020-08/fema488_mat_report_hurricane_charley_fl.pdf
FM-05 FEMA 549, Hurricane Katrina MAT. https://www.fema.gov/sites/default/files/documents/fema_549_hurricane_katrina_mat_report_2006.zip
FM-07 FEMA P-2023, Hurricane Irma MAT. https://www.fema.gov/sites/default/files/2020-07/mat-report_hurricane-irma_florida.pdf
FM-08 FEMA P-2077, Hurricane Michael MAT. https://www.fema.gov/sites/default/files/documents/fema_mat-report_hurricane-michael_florida.pdf
FM-11 FEMA P-2342, Hurricane Ian MAT. https://www.fema.gov/sites/default/files/documents/fema_rm-hurriance-ian-mat-report-12-2023.pdf
N-21 FEMA Marshall Fire MAT, structure-to-structure fire spread; full report P-2320. https://fema.gov/sites/default/files/documents/fema_rsl_marshall-mat-decreasing-risk-of-structure-to-structure-fire-spread-in-a-wildfire_042025.pdf ; https://www.fema.gov/sites/default/files/documents/fema_rsl_marshall-fire-mat-report-p-2320_042025.pdf
N-64 FEMA P-957, Snow Load Safety Guide. https://www.govinfo.gov/content/pkg/GOVPUB-HS5_100-PURL-gpo59955/pdf/GOVPUB-HS5_100-PURL-gpo59955.pdf

N-103 FEMA Marshall Fire MAT, Homeowner's Guide to Reducing Risk of Structure Ignition from Wildfire (rev. April 2025, DR-4634). https://fema.gov/sites/default/files/documents/fema_rsl_marshall-mat-homeowners-guide-to-reducing-risk-of-structure-ignition-from-wildfire_042025.pdf

7.5 USGS

N-01 USGS, "New USGS map shows where damaging earthquakes are most likely to occur in US" (Jan 16, 2024). https://www.usgs.gov/news/national-news-release/new-usgs-map-shows-where-damaging-earthquakes-are-most-likely-occur-us
M-26 USGS 2023 50-State National Seismic Hazard Model. https://www.usgs.gov/programs/earthquake-hazards/science/2023-50-state-long-term-national-seismic-hazard-model
N-02 Petersen et al., "The 2023 US 50-State National Seismic Hazard Model," Earthquake Spectra 40(1), 2024. https://par.nsf.gov/servlets/purl/10503874
N-09 USGS, M6.0 South Napa earthquake (2014). https://www.usgs.gov/programs/earthquake-hazards/science/m60-south-napa-california-earthquake-august-24-2014
N-70 USGS I-1940, swelling clays map. https://pubs.usgs.gov/publication/i1940
N-71 Jones & Holtz 1973, expansive soils damage estimate (TRID abstract). https://trid.trb.org/view/133235

7.11 Test videos (supporting links only; cite the reports first)

VID-01 "Ensayo de edificio sobre mesa sísmica", YouTube channel Salvador Ivorra (posted 5 Aug 2014). Eucentre, 7 Dec 2011; traces to [N-11, SCIP-32]. https://www.youtube.com/watch?v=p3Cf7hWkcYY
VID-02 "Prove sismiche su tavola vibrante di un modello a due piani realizzato con pannelli EMMEDUE®", EdiltecnicoTV (16 Apr 2009). ENEA Casaccia; see [SCIP-39]. https://www.youtube.com/watch?v=A3EPO3TAR_k
VID-03 "HORMI2 - TEST SISMICO PUCP", HORMI2 (posted 20 Nov 2013). PUCP, Lima; see [SCIP-40]. https://www.youtube.com/watch?v=As-f4gbIkY8
VID-04 "Seismic Tests r 2023", EMMEDUE SRL (21 Feb 2023). Manufacturer compilation. https://www.youtube.com/watch?v=-XwM88TjKU8
VID-05 "Test Sismico (ISRIM Università di Perugia) pannelli modulari emmedue", YouTube channel casasubito (about 2009). Not traced to a report. https://www.youtube.com/watch?v=qKEH6z6deUA

7.12 Exterior face: exposed foam, finishes, wildfire and hurricane observations (added 30 Oct 2026; URLs accessed 2026-10-30)

EV-01 ICC-ES ESR-2092, Nudura ICF (Tremco CPG), reissued Nov 2025, rev. Sept 2026: §2.0 and §4.3.3.1 (exterior covering required, fastened to the web flanges), §4.4.2.1 (rating from both sides only with gypsum or ≥1 in. concrete or masonry outside), §4.5 (Types I–IV with listed EIFS, 7/8-in. plaster or brick). https://icc-es.org/wp-content/uploads/report-directory/ESR-2092.pdf
EV-02 International Residential Code 2021, §R316 (foam plastic), §R608.4.3 (ICF exterior wall covering), §R703.9.1 (EIFS over concrete or masonry), §R301.2.1.2 (glazing protection); 2024 IRC §R303 (foam plastic, renumbered) and §R608.4.3; read on UpCodes (Texas and GSA adoptions). https://up.codes/viewer/texas/irc-2021/chapter/6/wall-construction ; https://up.codes/viewer/texas/irc-2021/chapter/3/building-planning ; https://up.codes/viewer/general-services-administration/irc-2024/chapter/3/building-planning
EV-03 International Building Code 2021, §2603.3, §2603.4, §2603.4.1.1, §2603.5, §2603.5.5 (NFPA 285 and Exception 2), §2603.5.7; read on UpCodes (Texas adoption). https://up.codes/viewer/texas/ibc-2021/chapter/26/plastic
EV-04 International Wildland-Urban Interface Code 2021, §503.2, §504.5, §505.5, §506 (Montana adoption), and 2024, §504.5 and §504.5.1 (GSA adoption), on UpCodes. https://up.codes/viewer/montana/iwuic-2021/chapter/5/special-building-construction-regulations ; https://up.codes/viewer/general-services-administration/iwuic-2024/chapter/5/special-building-construction-regulations
EV-05 California Residential Code 2022 §R337.7.3–R337.7.4 and Building Code 2022 §707A.3–707A.4.1; California Wildland-Urban Interface Code 2025 (Title 24 Part 7) §504.5, on UpCodes. https://up.codes/viewer/california/ca-building-code-2022/chapter/7A/sfm-materials-and-construction-methods-for-exterior-wildfire-exposure ; https://up.codes/viewer/california/ca-wuic-2025/chapter/5/special-building-construction-regulations
EV-06 CAL FIRE Office of the State Fire Marshal, Building Materials Listing Program, Listed Wildland-Urban Interface (WUI) Products Handbook, 2 Sept 2025 (Category 8140, SFM 12-7A-1; county-hosted copy). https://www.trinitycounty.org/DocumentCenter/View/3357/Wildland-Urban-Interface-Approved-Listed-Products-Handbook-
EV-07 A.J. Parker and D.A. Martin (Jensen Hughes), "Investigating ICF Wall Construction Meeting the Requirements of NFPA 285", NRMCA / ICFMA, June 2025 (industry-funded; tests at Intertek). https://www.nrmca.org/wp-content/uploads/NRMCA_ICFMA_NFPA_285_Report.pdf
EV-08 Wind Science and Engineering Research Center, Texas Tech University (with FAMU/FSU and UF), "A Summary Report on Debris Impact Resistance of Building Assemblies", submitted to NIST, August 2006, Appendix A, Tables A.1, A.4 and A.5 (ICF tests 30–33). https://www.depts.ttu.edu/nwi/research/DebrisImpact/Reports/TTU_Final_NIST_Report_numbered.pdf
EV-09 PCA / NAHB, "Concrete Homes Technology Brief No. 7: Concrete Homes Built-In Safety" (IS306), summarising PCA RP122 (industry-sponsored). https://www.nahb.org/-/media/NAHB/nahb-community/docs/councils/bsc/concrete-home-technology-briefs/IS306-concrete-homes-technology-brief-no7.pdf
EV-10 Sto Corp., "StoTherm ci – Impact resistance", sales sheet S902, Aug 2023 (ASTM E2486 classes). https://stocorp.com/wp-content/content/Systems_Marketing/Continuous%20Insulation%20Wall%20Systems%20(EIFS)/Sales%20Sheets/SS_Sto_Impact_Resistance_EN_web_S902.pdf
EV-11 Sika / Parex, "Procedures for Repairing EIFS", technical bulletin (rev. May 2025). https://usa.sika.com/dms/getdocument.get/6f28030d-3137-4c6c-aa1e-d9c69b194c64/us-parex-tech-bulletin-procedure-repairing-eifs.pdf
EV-12 D. Gorham, J. Willi and G. Horn (UL FSRI), "Residential Exterior Wall Reaction to Post-Flashover Compartment Fires", Fire and Materials 49(5):708–727, 2025, doi:10.1002/fam.3278; FSRI summary, 19 Mar 2025. https://doi.org/10.1002/fam.3278 ; https://fsri.org/research-update/journal-article-investigates-role-residential-siding-materials-spread-exterior
EV-13 FEMA P-2320, Mitigation Assessment Team Report: Marshall Fire (rev. April 2025), wall-construction observations and the rigid-foam research note. https://www.fema.gov/sites/default/files/documents/fema_rsl_marshall-fire-mat-report-p-2320_042025.pdf
EV-14 FEMA P-737, Builder's Guide to Construction in Wildfire Zones, December 2025, marked "pre-decisional draft" (copy hosted by Cal OES; cite with caution). https://www.caloes.ca.gov/wp-content/uploads/CWMP/2025-fema_building-science_builders-guide-construction-wildfire-zones_p-737.pdf
EV-15 A. Maranghides et al., WUI Structure/Parcel/Community Fire Hazard Mitigation Methodology, NIST TN 2205, 2022. https://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.2205.pdf
EV-16 A. Maranghides et al., A Case Study of the Camp Fire – Fire Progression Timeline, NIST TN 2135, Feb 2021 (interview table: "stucco survived but framing burned out"). https://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.2135.pdf
EV-17 IBHS, Early Insights: Lahaina Fire—2023, Aug 2023. https://firesafemarin.org/wp-content/uploads/IBHSEarlyInsights-LahainaFire.pdf
EV-18 B. Hajdukovič et al., large-scale fire tests of EPS ETICS facades, Fire Technology 53(1):173–209, 2017 (abstract record; window-plume fire, not wildfire). https://plus.cobiss.net/cobiss/si/en/bib/2240871
EV-19 L. Olsø, A. Haukø and B. Risholt (SINTEF), "Experimental study of fire exposed expanded polystyrene (EPS) insulation protected by selected coverings", Heliyon 10(4):e26309, Feb 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC10884846/
EV-20 Australian EPS ICF makers' bushfire claims (industry): Thermacell, BAL-29 under AS 1530.8.1 with acrylic render; ZEGO HomeFORM "up to BAL-29", with a different product for BAL-40 and flame zone. https://www.thermacell.com.au/the-blocks ; https://zego.com.au/products/insulated-concrete-forms/
EV-21 V. Salvia, "Recovering From the California Wildfires: How ICF is Playing a Role", ICF Builder Magazine, 1 Sept 2019 (trade press; Valley Fire 2015 outbuilding). https://icfmag.com/2019/09/recovering-from-the-california-wildfires-how-icf-is-playing-a-role/
EV-22 C. Ackley (Cal Poly student paper), republished as "A Case Study on How Insulated Concrete Forms Can Prevent Structure Loss During Wildfires", ICF Builder Magazine, 24 Aug 2021 (Camp Fire 2018); Nudura case study, Erlebach residence (Witch Creek 2007). Industry and trade press. https://icfmag.com/2021/08/a-case-study-on-how-insulated-concrete-forms-can-prevent-structure-loss-during-wildfires/ ; https://www.nudura.com/case-studies/erlebach-residence
EV-23 FEMA 489, Hurricane Ivan in Alabama and Florida, Building Performance Assessment, Aug 2005, Sec. 5.3.1 (EIFS and stucco failures, including over cast concrete). https://fema.gov/sites/default/files/2020-08/fema_489__hurricane_ivan_bpat_.pdf
EV-24 FEMA 548, Summary Report on Building Performance: Hurricane Katrina 2005, April 2006, Sec. 2.2.2.2 and 4.3.2. https://www.fema.gov/sites/default/files/2020-08/fema_548_summary_report_building_performance_hurricane_katrina_2005.pdf
EV-25 FEMA P-757, Hurricane Ike in Texas and Louisiana MAT, April 2009 (wall coverings; EIFS; stucco and lath penetrated by debris). https://www.fema.gov/sites/default/files/2020-08/fema757.pdf
EV-26 FEMA P-2022, Hurricane Harvey in Texas MAT, Feb 2019, Sec. 4.1.2 and 4.1.8 (vinyl loss; debris impacts on stucco and fiber cement). https://www.fema.gov/sites/default/files/documents/mat-report_hurricane-harvey-texas.pdf
EV-27 FEMA P-499, Home Builder's Guide to Coastal Construction (2010), Fact Sheets 1.8 (non-traditional systems: ICF, SIPs) and 5.3 (siding in high-wind regions). https://www.fema.gov/sites/default/files/2020-08/fema499_2010_edition.pdf
EV-28 IBHS, Hurricane Michael Wind Damage Assessment, Executive Summary (2019; 186 residences). https://ibhs.org/wp-content/uploads/IBHS-Hurricane-Michael-Executive-Summary.pdf
EV-29 D. Prevatt et al. (University of Florida), "Survey and Investigation of Buildings Damaged by Category-III, IV & V Hurricanes in FY 2018-2019 – Hurricane Michael", final report to Florida DBPR, 10 June 2019. https://www.floridabuilding.org/fbc/commission/FBC_0619/Prevatt-UF-Hurricane_Michael_Report.pdf
EV-30 D. Prevatt, D. Roueche and K. Gurley, "Hurricane Ian Building Damage Observation in FY 2022-2023", final report to the Florida Building Commission, 13 June 2023. https://www.floridabuilding.org/fbc/commission/FBC_0623/Prevatt-Hurricane_Ian_Building_Damage_Observation_in_FY_2022-2023_Final-v1.pdf
EV-31 "Mexico Beach Survivor", ICF Builder Magazine, 1 Sept 2019 (industry press), and S. Gibson, "A Bright Spot in the Ruins of Mexico Beach", Green Building Advisor, 16 Oct 2018. https://icfmag.com/2019/09/mexico-beach-survivor/ ; https://www.greenbuildingadvisor.com/article/bright-spot-ruins-mexico-beach
EV-32 Fox Blocks (Airlite Plastics), "A Guide to Selecting the Best Siding for Your ICF Wall" and FAQ "Does expanded polystyrene (EPS) present a serious fire hazard?" (manufacturer, undated). https://www.foxblocks.com/blog/icf-exterior-wall-finishes ; https://www.foxblocks.com/search/p16?query=6
EV-33 ICC-ES ESL-1453, Schnell Home sandwich panels (reissued Aug 2026): 1-h ASTM E119 wall, 3.15-in. EPS core, 1¾-in. concrete faces. https://icc-es.org/wp-content/uploads/report-directory/ESL-1453.pdf
EV-34 BRANZ Appraisal No. 750 (2023), EVG 3D Building System, 29 Nov 2023 (40–80 mm wythes; weatherproof coating required; impact resistance). https://www.branz.co.nz/assets/certified-products/750_2023.pdf
EV-35 A. Maranghides et al., NIST Outdoor Structure Separation Experiments (NOSSE) with Wind, NIST TN 2253, May 2023. https://doi.org/10.6028/NIST.TN.2253
EV-36 FEMA, Hurricane Michael Recovery Advisory 2, "Best Practices for Minimizing Wind and Water Infiltration Damage", June 2019. https://www.floridabuilding.org/fbc/commission/FBC_0220/HRAC/MichaelRA2_060719_508_FNALforposting.pdf
EV-37 Sika, "Stucco over CMU", technical bulletin (rev. May 2026). https://usa.sika.com/dms/getdocument.get/69f4e09e-54ac-4e3c-a691-d7c417247653/us-facades-tech-bulletin-stucco-over-cmu.pdf
EV-38 S. Mickley, "What is the most common finish used to clad a U.S. house?", AIBD, 2 Sept 2022 (Census Survey of Construction, 2021 starts). https://aibd.org/?p=24470
EV-39 Mark Anderson Design Build, "Building with the SCIP Construction Process" (builder testimonial, undated; Valley Fire 2015). https://markandersondesignbuild.com/fire-safe-building-technologies/
EV-40 Green Building Advisor Q&A, "Building With ICFs in a Fire-Prone Region", Sept 2020 (forum anecdote: EPS melted behind an intact synthetic stucco skin). https://www.greenbuildingadvisor.com/question/icf-house
EV-41 FEMA P-2020, Hurricanes Irma and Maria in Puerto Rico MAT, Oct 2018 (concrete-house performance; stucco coat cracking and spalling on a concrete hospital). https://www.fema.gov/sites/default/files/2020-07/mat-report_hurricane-irma-maria-puerto-rico_2.pdf

7.13 Termites, insurance and latent moisture (added 30 Oct 2026; URLs accessed 2026-10-30)

TM-01 Florida Building Code, Building, 8th ed. (2023), §2603.8 Protection against termites (UpCodes). https://up.codes/viewer/florida/fl-building-code-2023/chapter/26/plastic
TM-02 Florida Building Code, Building, 8th ed. (2023), §1816 Termite protection (soil treatment, baits, wood treatment) (UpCodes). https://up.codes/viewer/florida/fl-building-code-2023/chapter/18/soils-and-foundations
TM-03 Florida Building Code, Building, 8th ed. (2023), §1503.7 Protection against decay and termites (downspouts and condensate lines) (UpCodes). https://up.codes/viewer/florida/fl-building-code-2023/chapter/15/roof-assemblies-and-rooftop-structures
TM-04 Florida Building Code, Building, 8th ed. (2023), §1403.8 termite inspection clearance (UpCodes). https://up.codes/viewer/florida/fl-building-code-2023/chapter/14/exterior-walls
TM-05 Stronghold ICF System, Florida engineering evaluation report 0068-6-5-5939 (FL30543, FBC 2023), §2.3 citing FBC 2603.8 and FBC-R R318.4. https://www.floridabuilding.org/upload/PR_Tech_Docs/FL30543_R3_AE_0068-6-5_5939.pdf
TM-06 2025 California Residential Code §R305 (2024 IRC R305.1 and R305.4 model text; foam plastic protection and its three exceptions). https://gocodebook.com/library/us/ca/ca-residential-code/chapter-3-building-planning/section-r305-protection-against-subterranean-termites
TM-07 HUD Atlanta Homeownership Center, Appraisal and Property Requirements: TIP Zone information (Zone 1 "very heavy", Zone 2 "moderate to heavy"; masonry and concrete houses "no treatment needed"). https://archives.hud.gov/offices/hsg/sfh/ref/atlhoctip.pdf
TM-08 Oi, F. M. et al., Termite Prevention and Control, UF/IFAS Extension ENY-2044 (2020, rev. Dec 2023). https://edis.ifas.ufl.edu/publication/IN1277
TM-09 BuildBlock, Termite protection for below-grade applications (manufacturer technical data). https://buildblock.com/wp-content/uploads/Termite-Prevention-Strategies.pdf
TM-10 ICF Builder Magazine, "Termites and ICFs" (28 Sept 2017; trade press). https://icfmag.com/termites-and-icfs/
TM-11 Gromicko, N., "ICF Inspection and Termites", InterNACHI (undated; inspector's account). https://www.nachi.org/icf-termite-inspection.htm
TM-12 Georgia Department of Agriculture, press release on spray foam insulation and termite protection (2 Nov 2018). https://agr.georgia.gov/pr/improperly-installed-spray-foam-insulation-could-impact-ability-protect-termites
TM-13 Local Table R301.2(1) termite entries: Columbus, OH (CIC-09) https://www.columbus.gov/files/sharedassets/city/v/2/building-and-zoning/cics/cic-09-climatic-geographic-design-2026.pdf ; Spring Hill, TN https://springhilltn.org/DocumentCenter/View/394 ; Goodyear, AZ https://goodyearaz.gov/home/showpublisheddocument/24708/637342179022670000 ; Boise, ID https://codelibrary.amlegal.com/codes/boise_id/latest/boise/0-0-0-37350 ; Jefferson County, CO https://www.jeffco.us/DocumentCenter/View/1481/Climatic-and-Geographic-Design-Criteria-PDF ; Seattle, WA https://www.seattle.gov/documents/Departments/SDCI/Codes/SeattleResidentialCode/2018SRCChapter3.pdf ; Corona, CA https://codelibrary.amlegal.com/codes/corona/latest/corona_ca/0-0-0-75253 (Columbus and Spring Hill read in full; the others from search excerpts).
TM-14 Orkin, Termite Infestation Probability Zones map (pest-control industry; Hawaii Region I, Alaska Region IV). https://www.orkin.com/pests/termites/termite-infestation-probability-zones-map
IN-01 Texas Insurance Code §2006.002, Optional premium discount for insulating concrete form construction (FindLaw copy) https://codes.findlaw.com/tx/insurance-code/ins-sect-2006-002/ ; origin S.B. 581, 78th Legislature (2003), enrolled text https://capitol.texas.gov/tlodocs/78R/billtext/html/SB00581F.htm
IN-02 Family Building Society (UK), lending criteria: Property construction, insulating concrete form (last updated 7 Nov 2025). https://intermediaries.familybuildingsociety.co.uk/criteria/buy-to-let/details/bd4rw-property-construction-insulating-concrete-form-(icf)
LM-01 Building Science Corporation, High-R Foundation 13: Insulated Concrete Forms (ICF). https://buildingscience.com/documents/enclosures-that-work/high-r-value-wall-assemblies/high-r-foundation-13-insulated-concrete-forms-icf
LM-02 Iffa, E. and Tariku, F., "Hygrothermal performance assessment of ICF walls with different moisture control strategies and wall designs", Journal of Green Building 12(4):35–53 (2017). https://impact.ornl.gov/en/publications/hygrothermal-performance-assessment-of-icf-walls-with-different-m/
LM-03 ICF Builder Magazine, "Moisture and ICFs: The Facts" (14 May 2013; reports the 2002 PCA study by Gajda and VanGeem; trade press). https://icfmag.com/?p=9428
LM-04 Building America Building Science Advisor, "CMU Walls" (ORNL-hosted). https://hyperion.ornl.gov/pub/bsa/CMUWalls.pdf
LM-05 Lstiburek, J., BSI-029 "Stucco Woes: The Perfect Storm", Building Science Corporation. https://buildingscience.com/sites/default/files/document/bsi-029_the_perfect_storm_c.pdf

County shapes from the US Census via us-atlas. Shoreline counties from NOAA. Flood, hurricane and wildfire ratings from FEMA’s National Risk Index. Wildland-urban interface housing shares from the SILVIS Lab. Seismic design categories from the USGS ASCE 7-22 design maps.

The map

How the scoring works

Three axes, each 0 to 10

Evidence badge: shown beside each score, never inside it

Weighting

How the structure score is built

The whole house, not just the wall

The exterior face of the wall

Termites and latent moisture

Limitations

Summary

1. Region table

1.1 How the regions map

1.2 Regions

1.3 Overlay zones (these override the state colour inside the zone)

1.4 Machine-readable mapping

2. Methodology

2.1 Stage A: hazard performance

2.2 Stage B: deployability (the price and permitting axes)

2.3 What this method cannot do

2.4 Scoring for the map

3. Testing the case for SCIP: is it best for hurricane, earthquake and fire zones?

3.1 Hurricane: agree, with conditions

3.2 Tornado: not yet

3.3 Earthquake: partly agree (revised 28 Oct 2026)

3.4 Fire (wildfire): partly agree

3.5 Bottom line on the SCIP lean

3.6 Earthquake and wildfire regions: code approval versus demonstrated performance

3.7 The whole house: walls, roof and the connection between them

3.8 The exterior face: exposed foam, finishes and repair (new, 30 Oct 2026)

4. Evidence by hazard

4.1 Hurricane wind and windborne debris

4.2 Storm surge and flooding

4.3 Earthquake

4.4 Wildfire

4.5 Tornado

4.6 Hail

4.7 Snow, cold, freeze-thaw and energy

4.8 Termites, insurance, humidity and latent moisture (revised 30 Oct 2026)

4.9 Expansive soils

5. System verdicts

5.1 One paragraph per system

5.2 Capability matrix

6. Uncertainties and where experts disagree

7. Sources

7.1 SCIP approvals, evaluations and manufacturer tests

7.2 ICF, CMU, SIPs and other systems

7.3 Codes and standards (ICC, ASCE, Florida)

7.4 FEMA

7.5 USGS

7.6 NOAA

7.7 NIST

7.8 IBHS and FORTIFIED

7.9 Insurance, cost and policy

7.10 Research literature and field reports

7.11 Test videos (supporting links only; cite the reports first)

7.12 Exterior face: exposed foam, finishes, wildfire and hurricane observations (added 30 Oct 2026; URLs accessed 2026-10-30)

7.13 Termites, insurance and latent moisture (added 30 Oct 2026; URLs accessed 2026-10-30)

Related pages