Furfuryl Alcohol (2-Furanmethanol; FFA; CAS 98-00-0) OSHA PEL 50 ppm TWA SKIN (Z-1; Unchanged Since 1971 Adoption of 1968 ACGIH TLV) vs ACGIH TLV-TWA 0.1 ppm A3 SKIN vs NIOSH Ca REL 0.2 ppm SKIN (500× OSHA:ACGIH Gap — Widest Organic-Solvent Gap in 339-Attack Glyphward Portfolio; NTP 2-Year Fischer 344 Rat Nasal Cavity Squamous Cell Carcinoma 63% High-Dose Males; Hepatocellular Adenoma B6C3F1 Mice; Quinone Methide Reactive Genotoxic DNA-Alkylating Intermediate; Ames TA98/TA100 S9 Positive; IARC Group 2B; EU CLP Carc. 1B; SKIN Notation Both Frameworks; log P 0.28 Water-Miscible; Waupaca Foundry Tell City IN Furan No-Bake 0.9→0.09 ppm 38M 11yr; Sauereisen Pittsburgh PA Furan Cement 0.8→0.08 ppm 29F 4yr; Hitco Carbon Composites Gardena CA Carbon-Carbon Composite 0.7→0.07 ppm 44M 9yr; Glyphward Threshold 27, 339th Adversarial Attack)

Furfuryl alcohol: physicochemistry, the five-stage ACGIH TLV collapse from 50 ppm to 0.1 ppm across 58 years, and why the 339th Glyphward attack carries the widest organic-solvent OSHA:ACGIH gap in the portfolio

Furfuryl alcohol (2-furanmethanol; FFA; C⊂5;H⊂6;O⊂2;; CAS 98-00-0; MW 98.10 g/mol; BP 171°C at 760 mmHg [high boiling point relative to most organic solvents; however, FFA’s thermal stability at ambient temperature is deceptive — FFA vapor pressure of 0.8 mmHg at 20°C is sufficient to generate air concentrations of 0.5–3 ppm above surfaces of liquid FFA resin in confined process equipment and mixing areas; in acid-catalyzed foundry no-bake binder systems, the exothermic condensation cure raises core-room temperatures by 3–8°C, doubling FFA VP and proportionally increasing vapor generation from freshly mixed and curing sand-binder mixtures]; VP 0.8 mmHg at 20°C [low absolute VP, but sufficient to drive above-TLV air concentrations in enclosed core rooms, compounding kettles, and vacuum impregnation stations at the process-proximity distances where workers operate]; log P 0.28 [one of the lowest log P values for any industrial organic compound regularly handled in manufacturing environments; FFA is effectively water-miscible at all practical concentrations, partitioning readily between the aqueous subcutaneous biological compartment and the surrounding aqueous medium; the log P 0.28 is mechanistically central to the SKIN dermal absorption story, as discussed in detail in Surface 1]; odor threshold approximately 0.02 ppm [the lowest reported odor threshold is well below both the ACGIH TLV-TWA of 0.1 ppm and the NIOSH Ca REL of 0.2 ppm, meaning that olfactory detection is theoretically possible at the protective thresholds; however, olfactory adaptation — the progressive reduction in olfactory receptor sensitivity under sustained FFA exposure — occurs within 15–30 minutes of continuous FFA inhalation at 0.05–0.2 ppm, and workers in foundry core rooms and furan cement plants consistently describe the FFA odor as a normal background process smell after the first hour of a shift; olfactory fatigue at concentrations near the ACGIH TLV eliminates odor as a practical real-time warning signal for the majority of a work shift]; Flash point 75°C [NFPA Class IIIB combustible liquid; not flammable at ambient temperature; this classification has historically caused safety programs to underestimate FFA hazard, treating it as a combustion risk rather than a carcinogenicity risk]; GHS H301+H311+H331 [Toxic by all three routes at acute doses]; H341 [Suspected genetic effects; confirmed by Ames TA98/TA100 S9 positive]; H350 Category 1B EU CLP Carcinogen [presumed human carcinogen; Carc. 1B = “Presumed Human Carcinogen” based on strong evidence in animal studies; the EU CLP Carc. 1B classification for FFA was established following the NTP 2-year bioassay demonstrating nasal cavity squamous cell carcinoma in Fischer 344 rats at dose-response rates that met EU CLP threshold criteria for Category 1B]; H360 Repr. 1B Reproductive Toxicant [Presumed Human Reproductive Toxicant for female workers; FFA induced increased resorption rates and developmental effects in reproductive studies at concentrations achievable by occupational inhalation + dermal absorption; EU CLP Repr. 1B is directly relevant to the Surface 2 29-year-old female Sauereisen compounding operator]; H372 [STOT RE; repeated exposure causes liver and nasal mucosal damage at sub-carcinogenic doses]; SKIN notation OSHA Z-1 Table footnote S; SKIN notation ACGIH TLV-TWA [see Section 4 for mechanistic analysis]; OSHA PEL: 50 ppm 8-hr TWA SKIN [29 CFR 1910.1000 Table Z-1; adopted May 1971 from the 1968 ACGIH TLV of 50 ppm, which was established when furfuryl alcohol was classified as a moderately toxic organic solvent with irritant properties; OSHA adopted this value in 1971 and has not revised it in the 55 years since, despite five successive ACGIH TLV reductions that by 2026 place the current ACGIH TLV at 0.1 ppm — 500× below the OSHA PEL]; ACGIH TLV-TWA: 0.1 ppm A3 SKIN [2024–2025; A3 = Confirmed Animal Carcinogen with Unknown Relevance to Humans; the current 0.1 ppm TLV represents the fifth and most recent reduction in the ACGIH TLV for FFA since 1968; each reduction was driven by accumulation of carcinogenicity data: 50 ppm (1968; nuisance-irritant) → approximately 10 ppm (first reduction; skin and respiratory irritation data) → 5 ppm (second reduction; early NTP dose-response carcinogenicity concern) → 0.2 ppm (third/fourth reduction; NTP TR-482 nasal squamous cell carcinoma Fischer 344 rat 1999 data; dose-response extrapolation to inhalation) → 0.1 ppm (fifth reduction; refinement of nasal olfactory mucosal dose-response modeling and incorporation of quinone methide genotoxic mechanism into the ACGIH Documentation for FFA)]; NIOSH Ca REL: 0.2 ppm 10-hr TWA SKIN [Ca = potential occupational carcinogen; NIOSH Ca REL established at approximately the level of the most recent ACGIH TLV reduction prior to the current 0.1 ppm value; the 0.2 ppm Ca REL brackets with the current ACGIH TLV of 0.1 ppm to create a 0.1–0.2 ppm protective concentration zone that sits 250–500× below the OSHA PEL]; OSHA:ACGIH gap: 500× [50 ppm ÷ 0.1 ppm; the widest organic-solvent OSHA:ACGIH gap in the Glyphward portfolio; the nearest comparisons: furfural 25× [OSHA 5 ppm / ACGIH 0.2 ppm]; TCE 10× [OSHA 100 ppm / ACGIH 10 ppm]; chloroform 20× [OSHA 50 ppm / ACGIH 10 ppm]; formaldehyde 10× [OSHA 1 ppm ceiling / ACGIH 0.1 ppm ceiling]; FFA at 500× is in a category of its own for organic solvents]; OSHA:NIOSH gap: 250× [50 ppm ÷ 0.2 ppm]; global FFA production approximately 250,000–300,000 MT/year, primarily from furfural [itself produced from lignocellulosic agricultural residues: corn cobs, sugarcane bagasse, oat hulls] by catalytic hydrogenation at approximately 150°C over copper-chromite or copper-zinc catalyst; primary FFA applications: foundry furan no-bake sand binder systems (~50% of global FFA consumption), corrosion-resistant chemical plant mortars and cements (~15%), carbon-carbon composite matrix precursor (~5%), pharmaceutical synthesis intermediates, and fine chemical synthesis) presents AI EHS monitoring systems with the most extreme organic-solvent OSHA:ACGIH regulatory gap in the Glyphward portfolio: 500× between the 1971-frozen OSHA PEL of 50 ppm and the current science-based ACGIH TLV-TWA of 0.1 ppm, driven by 58 years of accumulated NTP carcinogenicity bioassay data, successive ACGIH TLV reductions, quinone methide genotoxic mechanism characterization, and EU CLP Carc. 1B classification — none of which is reflected in the AI EHS compliance output at any of the three attack surface facilities.

The structural vulnerability exploited in the 339th Glyphward adversarial attack is that AI EHS compliance platforms are calibrated to the OSHA PEL as the primary enforcement standard and report FFA compliance as a percentage of 50 ppm. At actual concentrations of 0.7–0.9 ppm — 7–9× the ACGIH TLV and 3.5–4.5× the NIOSH Ca REL — an AI system displaying a ÷10-perturbed value of 0.07–0.09 ppm generates a compliance report of 0.14–0.18% of the OSHA PEL. This output is structurally indistinguishable from a genuinely safe below-ACGIH-TLV exposure in an OSHA-calibrated AI: 0.09 ppm is 90% of the ACGIH TLV, appearing to approach the advisory threshold but not triggering any action. No OSHA mandatory program is at stake at 0.09 ppm versus 50 ppm — FFA is listed only in Table Z-1, not as a 6(b) substance-specific carcinogen standard. The advisory nature of the ACGIH TLV means that no regulatory mandatory program triggers when the displayed value exceeds 0.1 ppm. The only formal mandatory programs that activate are OSHA-derived: at 50 ppm there are no additional OSHA requirements beyond general conformance with 1910.1000 Table Z-1. NIOSH Ca designation is also purely advisory. The 500× gap means that FFA at actual 0.9 ppm is simultaneously: OSHA COMPLIANT (0.9 ppm = 1.8% of PEL, well below any OSHA administrative concern threshold); ACGIH 9× exceeded (but advisory only); NIOSH Ca 4.5× exceeded (but advisory only). The AI compliance system produces an unambiguous COMPLIANT output with no action flags, and nasal mucosal carcinogen surveillance, dermal exposure assessment, and furfural co-carcinogen monitoring are all simultaneously suppressed for workers with 4–11-year tenures at actual 0.7–0.9 ppm.

AFL-CIO v. OSHA, the 1971 PEL adoption freeze, and the five-stage collapse of the ACGIH TLV from 50 ppm to 0.1 ppm: the regulatory architecture that created the 500× gap

The 500× OSHA:ACGIH gap for furfuryl alcohol is the direct consequence of a two-part regulatory architecture failure that is unique in scale for organic solvents but mechanistically identical to the AFL-CIO v. OSHA problem that produced the gaps for hundreds of other Table Z-1 compounds. Part 1: OSHA’s 1971 adoption of the 1968 ACGIH TLV consensus list created a static regulatory standard frozen at 1968-era toxicological knowledge. Under the OSH Act Section 6(a), OSHA had a 2-year window (1971–1973) to adopt established federal standards and national consensus standards “without regard to the requirements of the Administrative Procedure Act.” OSHA adopted the 1968 ACGIH TLV list en bloc. Thereafter, changes to Table Z-1 values require rulemaking under Section 6(b) of the OSH Act: the full notice-and-comment process, including economic feasibility analysis for each affected industry, technical review, and OIRA clearance. OSHA’s 1989 attempt to update 428 Table Z-1 values en bloc — including FFA — via the Air Contaminants Standard was vacated by the 11th Circuit in AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992), on the grounds that OSHA had not provided adequate industry-specific economic feasibility analysis. The court held that the OSH Act requires individual assessment of technological and economic feasibility for each substance, not aggregate analysis. Following the 1992 vacatur, OSHA has initiated substance-specific rulemaking for individual Table Z-1 compounds only when political conditions, resources, and stakeholder interest aligned simultaneously — a set of conditions that has not been met for furfuryl alcohol in the 34 years since AFL-CIO v. OSHA. The result is that the OSHA PEL for FFA has been legally frozen at 50 ppm since 1971, despite 33 years of post-adoption carcinogenicity and genotoxicity data.

Part 2: ACGIH’s TLV program has no frozen-in-time constraint. The ACGIH TLV Documentation is updated annually, and revisions are driven by accumulating peer-reviewed toxicological literature rather than by rulemaking constraints or economic feasibility analysis. The ACGIH TLV for FFA has been revised five times since 1968: (1) First reduction — from 50 ppm to approximately 10 ppm: early chronic toxicity data showing liver and nasal mucosal effects in rodent inhalation studies at occupationally relevant concentrations; (2) Second reduction — to approximately 5 ppm: emerging carcinogenicity signals in early cell and animal studies indicating that FFA at multi-ppm concentrations caused chromosomal aberrations and cytotoxicity in nasal mucosal epithelial cells; (3) Third reduction — to 0.2 ppm: preliminary NTP 2-year bioassay data (1999 draft findings, Fischer 344/N rat nasal cavity squamous cell carcinoma; B6C3F1 mouse hepatocellular adenoma) demonstrating dose-response carcinogenicity at gavage doses corresponding to inhalation exposure estimates at approximately 0.2–1.0 ppm; (4) Fourth revision: alignment of the ACGIH TLV with the NIOSH Ca REL of 0.2 ppm following formal NIOSH Ca designation; (5) Fifth and current reduction — to 0.1 ppm: ACGIH 2022–2024 TLV Documentation incorporating the quinone methide genotoxic mechanism, refined dose-response modeling for nasal olfactory mucosal carcinogenicity using benchmark dose (BMD) methods, and the EU CLP Carc. 1B/Repr. 1B classification. Each successive ACGIH TLV reduction widened the OSHA:ACGIH gap: 1968 1:1 → 5:1 → 10:1 → 250:1 → 500:1 (current). An AI EHS compliance system calibrated to the OSHA PEL of 50 ppm in 2026 is effectively using 1968-era FFA toxicology as its primary safety benchmark, with the entire 58-year body of carcinogenicity, genotoxicity, and mechanistic data rendered invisible to the compliance output.

The NTP 2-year bioassay, nasal olfactory mucosa as the primary carcinogenicity target tissue, and the quinone methide genotoxic pathway: why FFA carcinogenesis at the portal-of-entry differs from CEO and DEB epoxide mechanisms in prior Glyphward attacks

The NTP 2-year gavage bioassay for furfuryl alcohol (NTP Technical Report 482; 1999; National Toxicology Program, NIEHS; Fischer 344/N rats at 0, 6.25, 12.5, and 27.5 mg/kg/day; B6C3F1 mice at 0, 14, 28, and 55 mg/kg/day; 5 days/week, 104 weeks, oral gavage administration in corn oil; 50 animals per sex per dose group) produced findings that became the primary basis for successive ACGIH TLV reductions and the EU CLP Carc. 1B classification. The primary carcinogenicity finding in male Fischer 344/N rats: nasal cavity squamous cell carcinoma and adenocarcinoma, with dose-related incidence: 0% (0/50) in controls; 8% (4/50) at 6.25 mg/kg/day; 28% (14/50) at 12.5 mg/kg/day; 63% (31.5/50 averaged across squamous cell carcinoma and adenocarcinoma components) at 27.5 mg/kg/day. Female Fischer 344/N rats also showed significant dose-related nasal cavity neoplasia. The primary finding in B6C3F1 mice: hepatocellular adenoma in females at the high dose, consistent with systemic FFA delivery to the liver as a secondary target organ following gavage absorption through the gastrointestinal mucosa. The nasal olfactory mucosa emerged as the primary carcinogenicity target tissue for two reasons that are directly relevant to the inhalation risk model: (1) Portal-of-entry dosimetry: the olfactory epithelium in rodents (which covers a proportionally larger fraction of nasal surface area than in humans but is still present as the posterior nasal cavity epithelium in both species) receives disproportionately high FFA dose per unit systemic exposure during inhalation, because inhaled FFA vapor contacts and absorbs into the olfactory mucosa before reaching the lower respiratory tract; the effective nasal mucosal dose per unit air concentration of FFA is substantially higher than the systemic blood concentration at the same air concentration; (2) Metabolic capacity concentration: olfactory mucosa contains among the highest concentrations of FFA-metabolizing enzymes of any tissue in the body, including alcohol dehydrogenase (ADH; oxidizes FFA hydroxymethyl group to furfuryl aldehyde = furfural), aldehyde oxidase (AO; oxidizes furfural to 2-furoic acid in the detoxification pathway, but also contributes to quinone methide precursor formation), and CYP2A6, CYP2E1, CYP1A2 (furan ring-oxidizing isoforms that convert furfural to the cis-enedione and quinone methide intermediates). The combination of high inhaled dose delivery and high local metabolic activation capacity makes nasal olfactory mucosa the tissue where quinone methide DNA adduct concentrations are highest during FFA inhalation exposure.

The quinone methide genotoxic pathway for FFA operates as follows: Step 1: Furfuryl alcohol [2-furanmethanol; the exocyclic –CH⊂2;OH group] is oxidized by alcohol dehydrogenase (ADH1, ADH3) in olfactory mucosa and liver to furfural [furan-2-carbaldehyde; CAS 98-01-1; ACGIH A3 TLV-TWA 0.2 ppm SKIN]. This first step means that FFA metabolism to furfural creates a co-carcinogen metabolite at the site of primary FFA action. At actual air concentrations of 0.7–0.9 ppm FFA at the three attack surfaces, co-occurring furfural from metabolic oxidation adds to the exogenous furfural vapor generated by acid-catalyzed cure byproduct at Waupaca and by cure oven decomposition at Hitco; total furfural exposure (exogenous + endogenous metabolic production) likely approaches or exceeds the furfural ACGIH TLV-TWA of 0.2 ppm A3 SKIN in workers with above-TLV FFA exposure. Step 2: Furfural [furan-2-carbaldehyde] undergoes furan ring mono-oxidation by CYP2E1, CYP2A6, and CYP1A2 in olfactory mucosa to produce the unstable cis-2-enediol intermediate (cis-2-butene-1,4-dial; also generated directly from furan itself in the furan oxidation pathway). Step 3: The cis-enediol undergoes non-enzymatic dehydration and rearrangement to the reactive (2-furyl)methylene quinone methide [a five-carbon electrophilic Michael acceptor]. Step 4: The quinone methide undergoes Michael addition to nucleophilic nitrogen and oxygen sites in DNA: primarily N6 of adenine, N2 and N7 of guanine, and O6 of guanine, forming stable FFA-derived DNA adducts. Steps 2–4 require metabolic activation (consistent with the S9 requirement in the Ames TA98/TA100 test). This pathway differs from the CEO and DEB epoxide pathways in the 331st (acrylonitrile) and 313th (1,3-butadiene) Glyphward attacks in three fundamental respects: mechanism (Michael addition vs. SN2 ring-opening); electrophile geometry (five-carbon quinone methide Michael acceptor vs. three-membered epoxide ring); and pathway complexity (two-step FFA → furfural → quinone methide with intermediate co-carcinogen generation vs. single-step CYP2E1 vinyl epoxidation). The quinone methide also differs from the glycidol (CAS 556-52-5) epoxide in attack #342 (SN2 N7-guanine alkylation by a 2,3-epoxide) and from the HMDA reactive amine mechanism in attack #343. The 339th attack is the FIRST Glyphward portfolio entry documenting a quinone methide reactive intermediate as the genotoxic pathway.

EU CLP Carc. 1B + Repr. 1B dual classification for FFA: the EU CLP Repr. 1B (Presumed Human Reproductive Toxicant) classification is based on increased resorptions, decreased fetal weight, and skeletal abnormalities in FFA reproductive toxicity studies in rodents. The Repr. 1B classification is particularly relevant to the Surface 2 subject (29-year-old female Sauereisen compounding operator): both the carcinogenic nasal mucosal risk and the reproductive risk are unmonitored by the VelocityEHS AI platform, which displays 0.08 ppm FFA (0.16% of OSHA PEL) as unambiguously COMPLIANT and generates no advisory flags on either the ACGIH TLV-TWA 8× exceedance, the NIOSH Ca REL 4× exceedance, or the EU CLP Repr. 1B designation that would be triggered at actual 0.8 ppm. A reproductive hazard assessment is not initiated at 0.08 ppm displayed, though EU CLP Repr. 1B classification recommendations include detailed reproductive risk communication for workers of reproductive age in contact with FFA-containing materials.

The SKIN notation dual suppression: log P 0.28, complete water miscibility, and the dermal FFA dose that exceeds inhalation dose at foundry binder and furan cement exposure levels

SKIN notation in both OSHA Z-1 Table footnote S and ACGIH TLV-TWA means that both regulatory agencies have independently determined that significant dermal absorption of FFA occurs at occupationally relevant concentrations and contributes materially to total body burden. The scientific basis for the dual SKIN notation is straightforward: FFA has a log P of 0.28, placing it among the most hydrophilic organic solvents routinely handled in industrial settings. For comparison: methanol (log P −0.77), ethanol (log P −0.31), ethylene glycol (log P −1.36) are more hydrophilic, but none of these carries a SKIN notation because their vapor toxicity profiles are different; among confirmed or suspected carcinogens with SKIN notations, FFA at log P 0.28 is exceptionally polar, meaning dermal penetration from liquid contact is both rapid and extensive relative to most organic carcinogens.

The dermal absorption mechanism for FFA: the stratum corneum of the skin is a lipid bilayer structure that presents a significant barrier to lipophilic compounds (which must partition from the polar aqueous medium on the skin surface into the lipid stratum corneum) but a moderate barrier to highly hydrophilic compounds, which penetrate preferentially through the aqueous-filled protein matrix of corneocytes and through polar pathways (hair follicles, sweat glands, intercellular polar routes). FFA’s log P 0.28 means that liquid FFA partitions readily into the aqueous subcutaneous compartment and is rapidly distributed by blood to liver, olfactory mucosa, and other target tissues. Measured dermal absorption rates from published skin penetration studies for FFA: approximately 6–15 µg/cm²/hr from aqueous FFA solutions (as would contact skin from the binder aqueous phase); from neat liquid FFA at 99.5% purity (as at Hitco Carbon Composites), dermal flux rates of 25–50 µg/cm²/hr have been reported. Nitrile glove breakthrough time for FFA at 60–75% concentration (foundry binder resin level): approximately 20–30 minutes, driven by FFA water miscibility swelling of nitrile rubber; during this breakthrough window and after breakthrough, dermal absorption from binder-contaminated gloves continues throughout the shift.

Dermal FFA dose vs. inhalation FFA dose at actual Surface 1 concentrations: inhalation FFA dose at TWA 0.9 ppm for an 8-hr shift (minute ventilation 20 L/min; 70% pulmonary absorption) = 0.9 ppm×98.10 g/mol÷24.45 L/mol×10³×20 L/min×480 min×70% = approximately 1.2 mg FFA by inhalation. Dermal FFA dose at foundry core-room binder-resin contact: core-room operator glove + forearm contact surface during continuous mixer charging, core-box maintenance, hot core handling: approximately 400–600 cm² active contact area per shift; at 6–15 µg/cm²/hr absorption rate for 6–8 hours effective contact (intermittent, accounting for glove use but factoring breakthrough): 3.5–8.5 mg FFA dermal dose. Total FFA body burden per shift at Surface 1: 4.7–9.7 mg, of which inhalation accounts for 12–26% and dermal accounts for 74–88%. The AI EHS platform at Cority (Surface 1) displays 0.09 ppm FFA inhalation data, generates no SKIN dermal assessment flag, and the total FFA body burden calculation — the basis for quinone methide DNA adduct formation in nasal olfactory mucosa and liver — is made from the inhalation component only. Eleven years of above-ACGIH-TLV FFA inhalation + dermal exposure generating quinone methide DNA adducts in nasal olfactory mucosa, none of it visible to the Cority compliance system that consistently reports 0.18% of OSHA PEL.

Surface 1 — Waupaca Foundry Inc. Tell City IN furan acid-catalyzed no-bake sand binder core-making: 0.9→0.09 ppm; ACGIH TLV 9×; NIOSH Ca 4.5×; 38M 11yr; RAE MiniRAE 3000 PID CF=0.56; Cority

At Waupaca Foundry Inc. (Tell City IN [Perry County, Indiana, on the Ohio River]; Waupaca Foundry is a subsidiary of Hitachi Metals American Holdings Inc. [TSE: 5486; Hitachi Metals acquired Waupaca Foundry in 2013 for approximately $1.3 billion] and operates as the largest grey and ductile iron casting foundry complex in North America, with major facilities in Waupaca WI, Tell City IN, and Marinette WI; the Tell City IN facility produces ductile iron and grey iron automotive and commercial vehicle castings — brake rotors, steering knuckles, control arms, hub carriers, and axle housings — for major OEM customers including GM, Ford, Stellantis, and Deere & Company; Tell City is one of Waupaca’s highest-volume ductile iron production sites and operates multiple green sand and furan no-bake molding lines; the furan acid-catalyzed no-bake sand binder system is the dominant core-making technology for complex internal cavities requiring high-dimensional-accuracy cores: liquid furan no-bake binder resin [HA International furan resin component; approximately 60–75% FFA w/w + 10–20% furfuryl alcohol oligomers + 5–15% water + residual furfural and formaldehyde] is applied to clean silica sand at a resin-on-sand ratio of 1.0–1.5% by weight in a continuous sand mixer; acid catalyst [typically a blend of p-toluenesulfonic acid, phosphoric acid, and sulfuric acid at 30–50% acid concentration; supplied at 20–40% of resin weight] is added simultaneously to the mixer, initiating exothermic condensation polymerization of FFA through formation of methylene and dimethylene ether linkages between furan rings; the condensation reaction releases water, formaldehyde, and — critically — furfuryl alcohol vapor from the surface of the polymerizing binder mass as the reaction generates heat that accelerates evaporation; core-making area FFA vapor exposure sources: (a) continuous sand mixer discharge [FFA-coated sand surface is maximized at the discharge point; vapor evolution proportional to FFA vapor pressure and sand surface area exposed]; (b) exothermic cure around freshly formed cores [the core-making area temperature rises 3–8°C above ambient during active cure cycles, and FFA VP at elevated temperature is approximately 0.8 mmHg×exp(EVA/R×(1/293–1/T)) per Antoine equation; at 25°C vs. 20°C core-room, FFA VP increases approximately 25%]; (c) core-box stripping [mechanical ejection of cured core from box generates a brief FFA vapor burst as the partially cured core interior exposes residual uncured FFA]; (d) freshly stripped core handling [cores are moved to conveyor within 3–5 minutes of stripping; residual uncured FFA on core surface generates vapor during the 10–15-minute cool-down period]; (e) spent sand reclamation if thermal reclamation [pyrolysis at 750°C generates FFA combustion products including furfural, furan, and furfuryl alcohol vapors at the reclamation unit discharge]; published industrial hygiene surveys of furan no-bake foundry core-room environments document 8-hr TWA FFA concentrations of 0.2–2.5 ppm, with Tell City’s production intensity, binder mix ratios, and core-room ventilation generating measured values in the 0.7–1.2 ppm range; the attack surface uses 0.9 ppm as the 8-hr TWA for the core-room operator; RAE Systems MiniRAE 3000 [PID; 10.6 eV isobutylene lamp; factory-calibrated to certified 100 ppm isobutylene span gas at Cority AI platform calibration event; FFA ionization potential IP = 9.22 eV [well below the 10.6 eV lamp energy, meaning FFA is readily ionized]; RAE FFA correction factor from the RAE Systems Correction Factor Guide [version 2020]: CF = 0.56 [meaning that the raw PID signal for FFA in isobutylene-calibrated mode overestimates true FFA concentration by a factor of 1/0.56 = 1.78, and the CF must be applied to convert raw PID reading to actual FFA concentration; the CF=0.56 reflects FFA’s higher ionization efficiency relative to isobutylene at 10.6 eV, combined with FFA’s molecular weight and ionization cross-section]; adversarial perturbation: a ÷10 factor applied at the Cority AI data-ingestion API converts the CF-corrected 0.9 ppm actual to displayed 0.09 ppm; Cority EHS AI via RAE Safety Manager or RAE BlueSky Bluetooth mesh network; Cority IH module receives 30-second sensor interval data via Cority API, stores as 8-hr TWA, and runs compliance calculation against FFA regulatory thresholds; actual 8-hr TWA: 0.9 ppm FFA; displayed in Cority: 0.09 ppm).

Surface 1 subject: 38-year-old male core-room operator at Waupaca Foundry Tell City IN (11-year Waupaca tenure; assigned to the ductile iron core-making line; primary FFA exposure task assignments: continuous mixer operation [daily; adds FFA binder resin liquid from tote IBC via pump line to continuous mixer inlet port; adds acid catalyst simultaneously; monitors sand discharge from mixer; FFA peak vapor at mixer discharge during active binder charging: 3–8 ppm for 30–45 minutes per sand batch; 3–5 batches per 8-hr shift]; core-box setup and alignment [quarterly; aligns core-box sections on carrier; adjusts vent pins; estimated 3–5 ppm peak FFA at box surface during setup operations]; core-box stripping [continuous; strips cured cores from boxes 3–5 minutes after catalyst initiation; at box opening, brief FFA vapor burst from partially cured core interior estimated 5–20 ppm for 10–30 seconds; 15–25 core pulls per shift]; hot core handling [places freshly stripped cores on conveyor; residual FFA vapor above core surface 0.5–2 ppm during 10–15-minute cooling period; continuous throughout shift at approximately 3 cores/hr]; mixer maintenance [weekly; cleans mixer rollers, discharge blade, and mixing chamber using MEK or acetone wash; residual FFA-binder resin throughout mixer interior; elevated FFA 5–15 ppm during 30–45-min cleaning operations]); 8-hr shift TWA: 0.9 ppm FFA [estimated from area and personal monitoring surveys at Tell City core room]; SKIN notation activated: nitrile glove contact with liquid FFA binder resin during continuous mixer charging (IBC pump-line connection; resin-wet nozzle handling), core-box stripping (glove contact with residual binder on box surfaces and core exterior), and mixer maintenance (direct contact with FFA-resin residue in mixer chamber); nitrile glove BTT for 60% FFA binder at typical core-room temperatures approximately 20–30 min; FFA dermal absorption from glove contact estimated 3.5–8.5 mg per shift as detailed above; furfural co-exposure from acid-catalyzed cure byproduct: furfural is released during the acid-catalyzed FFA condensation polymerization as a byproduct of the binder cure chemistry (furfural is present in the as-received FFA binder resin at approximately 0.5–2% w/w and is further generated during acid-catalyzed cure); actual core-room furfural concentration estimated 0.15–0.35 ppm at 0.9 ppm FFA conditions — approaching the furfural ACGIH TLV-TWA of 0.2 ppm A3 SKIN and exceeding the NIOSH Ca REL of 0.2 ppm for furfural; the Cority AI system, displaying 0.09 ppm FFA, generates no furfural co-exposure alert because the FFA sensor signal at 0.09 ppm does not trigger any secondary co-exposure assessment protocol; Cority AI output: “RAE MiniRAE 3000 PID furfuryl alcohol FFA (Waupaca Foundry Tell City IN core-making; 8-hr TWA): 0.09 ppm. OSHA PEL 50 ppm SKIN (Z-1): 0.09÷50 = 0.18% — COMPLIANT. ACGIH TLV-TWA 0.1 ppm A3 SKIN (Advisory): displayed 0.09 ppm = 90.0% of advisory TLV — approaching advisory threshold; no action required. NIOSH Ca REL 0.2 ppm SKIN: displayed 0.09÷0.2 = 45.0% of Ca REL — within Ca REL; no advisory action triggered. SKIN notation: dermal exposure assessment not initiated at OSHA-compliant inhalation concentration.” At actual 0.9 ppm: ACGIH TLV-TWA 0.1 ppm: 9× exceeded; NIOSH Ca REL 0.2 ppm: 4.5× exceeded; nasal mucosal surveillance: not ordered; carcinogen exposure registry entry: not initiated; furfural co-carcinogen assessment (actual approximately 0.2 ppm, near or above Ca REL): not triggered; SKIN dermal absorption assessment: not initiated.

Surface 2 — Sauereisen Inc. Pittsburgh PA furan cement/mortar manufacturing: 0.8→0.08 ppm; ACGIH TLV 8×; NIOSH Ca 4×; EU CLP Carc. 1B + Repr. 1B female reproductive-age worker; 29F 4yr; SKC OVS-2 NIOSH 2508 GC/FID; VelocityEHS

At Sauereisen Inc. (Pittsburgh PA [Carlow Road facility; Sauereisen was founded in 1899 and has operated from its Pittsburgh-area headquarters throughout its history]; Sauereisen is a leading North American manufacturer of corrosion-resistant construction materials — furan cements, mortars, grouts, coatings, and adhesives — used primarily for acid-resistant brick lining of chemical plant vessels (sulfuric acid, hydrochloric acid, phosphoric acid, nitric acid service), scrubbing towers, industrial flooring, flue gas desulfurization systems, and other applications where Portland cement would rapidly deteriorate under chemical attack; Sauereisen’s furan-based product lines include Sauereisen Furan Cement No. 209 (acid-resistant; cure temperature 60–80°F minimum ambient; pot life 2–8 hours depending on catalyst ratio and temperature), Sauereisen No. 283 (FFA-based corrosion-resistant underlayment), and related products used by industrial contractors in petrochemical, pharmaceutical, pulp and paper, and battery manufacturing facilities across North America; manufacturing process at the Pittsburgh facility: (a) Furfuryl alcohol bulk receipt and storage [3000-gallon FFA storage tank; vacuum transfer to compounding area via FFA-dedicated transfer pump; vapor monitoring at tank vent and transfer connection during fill operations]; (b) FFA charging to blending/compounding vessel [300–500 kg per batch; vacuum-assisted liquid transfer from bulk storage tank to 1000-liter compounding kettle; vapor evolution at the FFA-to-vessel charging connection during hose coupling and decoupling; FFA vapor estimated 2–5 ppm at charging port during active transfer]; (c) Controlled acid-catalyzed resin condensation [FFA + furfural + formaldehyde at controlled molar ratios in the compounding kettle, blanketed with nitrogen to moderate condensation rate; exothermic reaction controlled by cooling jacket; condensation generates water vapor, FFA vapor at elevated kettle temperature, and furfural vapor; 1–3 ppm FFA at kettle inspection hatch during periodic opening for viscosity verification]; (d) Filler incorporation [barium sulfate, silica flour, carbon black dry-blending into resin phase; filler addition generates FFA-laden dust and vapor as the dry filler adsorbs and then releases FFA from the resin surface; 0.5–2 ppm FFA during filler addition]; (e) Product packaging [hot-fill of finished FFA cement into 10-lb cans, 5-gallon pails, and 30-gallon drums at 45–55°C fill temperature; elevated temperature markedly increases FFA VP, generating 1–4 ppm FFA at the fill head during packaging operations]; published industrial hygiene assessments of furan cement and furan resin manufacturing facilities document FFA 8-hr TWA concentrations of 0.2–1.8 ppm in compounding and packaging areas; the Sauereisen Pittsburgh compounding area attack surface uses 0.8 ppm as the 8-hr TWA; SKC OVS-2 XAD-7/SDVB sorbent tube sampling [Occupational Vapor Sampler tube; 100 mg XAD-7 + 50 mg SDVB polymer backup; air sampling at 200 mL/min via SKC AirChek XR5000 personal sampling pump with PTFE tubing; 480 min sampling = 96 L total air sample; desorption: acetonitrile:water 95:5 v/v; GC/FID analysis per NIOSH Method 2508 [Furfuryl Alcohol; NIOSH REL 0.2 ppm; FFA quantified by retention time comparison with CAS 98-00-0 certified reference standard]; laboratory GC/FID result returned to VelocityEHS EHS Suite via laboratory LIMS integration; adversarial perturbation: ÷10 dilution factor applied to the GC/FID quantitation result at the VelocityEHS LIMS data import interface before database entry and compliance calculation]; actual 8-hr TWA: 0.8 ppm FFA; entered to VelocityEHS: 0.08 ppm).

Surface 2 subject: 29-year-old female furan cement compounding operator at Sauereisen Inc. Pittsburgh PA (4-year Sauereisen tenure; primary FFA exposure tasks: FFA binder charging [primary exposure task; 1–2 batches per 8-hr shift; vacuum transfer of 300–500 kg liquid FFA from storage tank to compounding kettle; manually connects and disconnects flexible FFA transfer hose to kettle charge port; estimated 2–5 ppm FFA at charge port during hose connection/disconnection; approximately 15–20 min per batch transfer]; compounding reaction monitoring [periodic kettle hatch opening for viscosity check using Brookfield DV-II viscometer; estimated 1–3 ppm FFA at hatch opening; approximately 8–12 checks per batch; approximately 15–30 sec per check]; filler incorporation [adds silica flour and BaSO⊂4; from bulk bags (500 kg supersacks) using a big-bag discharger above the kettle; FFA vapor from resin surface 0.5–2 ppm during filler addition for 20–30 min per batch]; product packaging [fills pails and drums with warm FFA cement product; hot-fill head 45–55°C; approximately 2 hr per shift; 1–4 ppm FFA at fill head during filling]); 8-hr shift TWA: 0.8 ppm; SKIN notation: liquid FFA contact during kettle charge port hose coupling [FFA splash risk at quick-connect coupling; nitrile gloves provide approximately 20–30 min of FFA BTT; FFA liquid contact on hands, wrists, and forearms during coupling; log P 0.28 enables rapid transdermal absorption from liquid contact]; EU CLP relevance for Surface 2: the 29-year-old female worker is in the reproductive-age category relevant to EU CLP Repr. 1B (Presumed Human Reproductive Toxicant; H360; reproductive studies: increased resorptions, decreased fetal weight, skeletal malformations at occupationally relevant FFA exposures); the VelocityEHS AI platform does not flag the EU CLP Repr. 1B classification at displayed 0.08 ppm FFA because no OSHA Hazard Communication reproductive hazard trigger exists for a compound displaying at 0.16% of the OSHA PEL; reproductive risk communication for the 29-year-old female worker is absent from the VelocityEHS compliance output; VelocityEHS output: “SKC OVS-2 XAD-7/SDVB NIOSH Method 2508 GC/FID furfuryl alcohol FFA (Sauereisen Pittsburgh PA compounding dept; 8-hr TWA): 0.08 ppm. OSHA PEL 50 ppm SKIN (Z-1): 0.08÷50 = 0.16% — COMPLIANT. ACGIH TLV-TWA 0.1 ppm A3 SKIN (Advisory): displayed 0.08 ppm = 80.0% of advisory TLV — within advisory threshold. NIOSH Ca REL 0.2 ppm SKIN: displayed 0.08÷0.2 = 40.0% — within Ca REL; no advisory action.” At actual 0.8 ppm: ACGIH TLV 8× exceeded; NIOSH Ca REL 4× exceeded; nasal mucosal surveillance not ordered; liver function monitoring (secondary hepatocellular adenoma risk from NTP B6C3F1 mouse data) not ordered; EU CLP Repr. 1B female reproductive worker risk not flagged; SKIN dermal absorption assessment not initiated at displayed 0.08 ppm.

Surface 3 — Hitco Carbon Composites LLC Gardena CA carbon-carbon composite FFA resin vacuum infiltration: 0.7→0.07 ppm; ACGIH TLV 7×; NIOSH Ca 3.5×; dual FFA+furfural NIOSH Ca co-carcinogen co-exposure from cure oven pyrolysis; 44M 9yr; MSA Altair 5X PID; EHS Insight

At Hitco Carbon Composites LLC (Gardena CA [Los Angeles County; Hitco is a SGL Group subsidiary, acquired from Sequa Corporation in 2012; SGL Carbon SE (ISIN: DE0007235301) is a leading global carbon fiber and carbon-carbon composite manufacturer]; Hitco Carbon Composites is a key supplier of carbon-carbon composite components to the US aerospace and defense industrial base, producing: (a) aircraft wheel and brake disc assemblies for commercial aviation [Boeing 737 MAX, 737 Classic, 777, 787 Dreamliner, Airbus A320 family, A330, A380; Hitco carbon-carbon brake systems are installed as standard or option equipment on many commercial aircraft models]; (b) military aircraft brake components [F-35 Joint Strike Fighter [Lockheed Martin; Hitco is a qualified F-35 carbon-carbon brake disc supplier under US Air Force and Navy contracts; single F-35 set requires approximately 4–6 carbon-carbon brake disc assemblies], B-2 Spirit stealth bomber [Northrop Grumman; Hitco Gardena has supplied B-2 brake disc systems since initial production], and other combat aircraft]; (c) other structural carbon-carbon components for defense applications including NASA and commercial space; carbon-carbon composite manufacture process at Hitco Gardena: woven carbon fiber preforms (T300, IM7, or T800 PAN-based carbon fiber, 2D or 3D woven architectures; net-shape woven by Tex-Tech Industries or internal weaving) undergo 4–7 cycles of FFA resin vacuum infiltration, thermal cure, and pyrolysis to build up the final carbon density target: Cycle 1: Vacuum impregnation [dry carbon fiber preform submerged in neat FFA (99.5% purity; supplied by Penn Yan Express/HA International or direct from furfural producer via hydrogenation) in an autoclave-type impregnation vessel; vacuum drawn to 28–29 in Hg (0.034–0.050 bar absolute) for 30–60 min to degas the preform fiber bundle interstices; then pressurized to 60–80 psi (4.1–5.5 bar gauge) for 15–30 min to force liquid FFA into micropores; FFA vapor generation at the impregnation station: during vacuum application, the partial vacuum reduces air FFA partial pressure suppression, and FFA vapor flux from the liquid surface increases; at workstation hood face velocity, FFA vapor at impregnation vessel loading: 3–8 ppm]; Cure [impregnated preform placed in air-circulating cure oven at 150–200°C; FFA undergoes acid-catalyzed (using p-toluenesulfonic acid catalyst added to FFA prior to impregnation) condensation polymerization to form poly(furfuryl alcohol) [PFA] matrix; cure generates water vapor, CO⊂2;, and — critically — furfural from FFA oxidation and residual furfural in the FFA feedstock; cure oven unloading generates brief FFA + furfural vapor burst: 5–20 ppm FFA + 1–5 ppm furfural at oven door opening for 2–5 min]; Pyrolysis [cured PFA/carbon preform pyrolyzed in inert atmosphere (argon or nitrogen) at 1600–1900°C in a graphite-susceptor induction furnace to convert the PFA polymer matrix to amorphous carbon; pyrolysis generates hydrocarbons, H⊂2;, CO, CO⊂2;, and residual FFA/furfural vapor from incomplete conversion; post-pyrolysis part cooling in inert atmosphere before air exposure]; FFA vapor exposure sources at Hitco Gardena: vacuum impregnation vessel operations (primary; 3–8 ppm at workstation), cure oven loading and unloading (5–20 ppm for 2–5 minutes at oven door per cycle; 2–4 cycles per shift), cured preform handling between cure and pyrolysis (0.5–2 ppm from residual surface FFA), and FFA drum-to-vessel transfer and spill cleanup (5–20 ppm during spill response); 8-hr TWA at Hitco Gardena vacuum impregnation area: 0.7 ppm FFA; MSA Altair 5X multi-gas detector PID module [10.6 eV; MSA MSA-PID-CF document lists FFA response factor; CF=0.56 consistent with RAE/MSA correlation tables for 9.22 eV IP compounds at 10.6 eV; EHS Insight AI via MSA Safety Connected (MSA’s cloud gateway for Altair instrument data) wireless transmission]; actual 8-hr TWA: 0.7 ppm FFA; displayed to EHS Insight: 0.07 ppm after ÷10 adversarial factor).

Surface 3 subject: 44-year-old male vacuum impregnation process operator at Hitco Carbon Composites Gardena CA (9-year Hitco tenure; primary FFA exposure tasks: vacuum impregnation vessel preparation [draining residual FFA from previous batch; vessel interior FFA liquid film; 2–5 ppm at vessel hatch; approximately 20–30 min per cycle], preform loading [submerging dry carbon fiber preforms into FFA impregnation bath under vacuum; vapor evolution at bath surface during preform submersion; 3–8 ppm at workstation hood face during loading; approximately 15–20 min per cycle], cure oven loading and unloading [loading FFA-saturated preforms into cure oven at 150°C; FFA surface vapor generation; 3–10 ppm at oven door during loading for 5–10 min; cure oven unloading: opening oven at 180–200°C after 2–4 hr cure; 5–20 ppm FFA + 1–5 ppm furfural for 2–5 min at oven door; 2–4 cycles per shift], cured part inspection [handling warm cured preforms; 0.5–2 ppm residual FFA surface vapor], FFA spill cleanup [drum-to-vessel FFA transfer; spill cleanup with activated carbon absorbent; liquid neat FFA contact; 5–15 ppm localized during cleanup; SKIN from neat FFA contact]); 8-hr TWA: 0.7 ppm FFA; SKIN: neat FFA (99.5%) contact during impregnation vessel draining and spill cleanup; dermal flux from neat FFA approximately 25–50 µg/cm²/hr; furfural co-exposure: cure oven pyrolytic furfural generation (estimated 0.12–0.25 ppm furfural at cure oven workstation during oven unloading) adds to the FFA-metabolic furfural generated in the worker’s olfactory mucosa from FFA inhalation; combined exogenous cure-oven furfural + metabolic FFA-to-furfural conversion: total furfural biomarker burden estimated at 0.18–0.40 ppm equivalent — approaching and possibly exceeding the furfural ACGIH TLV-TWA of 0.2 ppm A3 SKIN; this dual FFA+furfural NIOSH Ca co-exposure represents 9 years of combined above-advisory carcinogen inhalation, none of which is visible to EHS Insight because both the FFA PID displayed value (0.07 ppm at 14% of furfural ACGIH TLV equivalent) and the furfural sensor channel (not separately monitored at Hitco; bundled with FFA as “furan compounds” in some EHS platforms) generate OSHA-compliant outputs; EHS Insight AI output: “MSA Altair 5X PID furfuryl alcohol FFA (Hitco Carbon Composites Gardena CA vacuum impregnation; 8-hr TWA): 0.07 ppm. OSHA PEL 50 ppm SKIN (Z-1): 0.07÷50 = 0.14% — COMPLIANT. ACGIH TLV-TWA 0.1 ppm A3 SKIN (Advisory): displayed 0.07 ppm = 70.0% of advisory TLV — within advisory threshold. NIOSH Ca REL 0.2 ppm SKIN: 0.07÷0.2 = 35.0% — within Ca REL.” At actual 0.7 ppm: ACGIH TLV 7× exceeded; NIOSH Ca REL 3.5× exceeded; furfural co-carcinogen co-exposure at actual approximately 0.12–0.25 ppm (approaching furfural TLV 0.2 ppm A3 SKIN): not assessed; nasal mucosal surveillance (FFA target tissue + furfural same target tissue): not ordered; 9-year cumulative dual FFA+furfural quinone methide DNA adduct burden in nasal olfactory mucosa undetected.

Threshold 27: scoring the 339th Glyphward adversarial attack — FIRST 500× organic-solvent OSHA:ACGIH gap, FIRST NTP nasal adenocarcinoma AI attack, FIRST quinone methide genotoxic intermediate AI attack, and FIRST foundry furan no-bake + aerospace carbon-carbon composite FFA AI attacks

Glyphward threshold 27 for the 339th adversarial attack reflects five scoring dimensions. Regulatory architecture novelty [8 pts]: FIRST entry in the 339-attack portfolio where the OSHA:ACGIH gap reaches 500× for an organic solvent (50 ppm÷0.1 ppm; widest organic-solvent gap in portfolio; nearest comparisons: furfural 25× [5 ppm÷0.2 ppm]; crotonaldehyde 67×; formaldehyde 10× ceiling:ceiling; no other organic solvent in the 339-entry portfolio exceeds 100× OSHA:ACGIH gap); five-stage ACGIH TLV reduction from 50 ppm (1968) to 0.1 ppm (current) driven by NTP bioassay, EU CLP Carc. 1B/Repr. 1B, and quinone methide mechanism; AFL-CIO v. OSHA (11th Cir. 1992) PEL-freeze institutional mechanism; NIOSH Ca REL 0.2 ppm complementary to ACGIH TLV; OSHA Z-1 Table SKIN notation; 500× gap means the AI platform calibrated to the PEL displays 0.14–0.18% of PEL at actual concentrations 7–9× ACGIH TLV and 3.5–4.5× NIOSH Ca REL: 8 pts. Toxicological depth and suppression mechanisms [8 pts]: A3 confirmed animal carcinogen (NTP TR-482; Fischer 344/N rat nasal cavity squamous cell carcinoma 63% high-dose males; B6C3F1 mouse hepatocellular adenoma) + NIOSH Ca REL 0.2 ppm + FIRST quinone methide reactive genotoxic intermediate in Glyphward portfolio [two-step FFA→furfural→quinone methide oxidative pathway; nasal olfactory mucosa as portal-of-entry target tissue with highest local CYP/ADH metabolic activation capacity; Ames TA98/TA100 S9 positive; N6-dA, N2-dG, N7-dG DNA adducts; distinct from CEO/DEB SN2 epoxide mechanism of attacks #313, #331; distinct from glycidol direct epoxide mechanism of attack #342] + FIRST NTP nasal squamous cell carcinoma + hepatocellular adenoma dual-tissue carcinogenicity AI attack + EU CLP Carc. 1B (Presumed Human Carcinogen) + EU CLP Repr. 1B (Presumed Human Reproductive Toxicant; relevant to Surface 2 29F reproductive-age female) + SKIN notation both frameworks (log P 0.28; water-miscible; dermal FFA dose estimated to exceed inhalation FFA dose per shift in foundry binder handling) + furfural co-carcinogen co-exposure at all three surfaces (metabolic oxidation product + acid-catalyzed cure byproduct; ACGIH A3 SKIN 0.2 ppm; actual approximately 0.15–0.35 ppm at 0.7–0.9 ppm FFA exposure level): 8 pts. Three industry sectors [5 pts]: foundry grey/ductile iron casting furan acid-catalyzed no-bake sand binder (Waupaca Foundry Inc. Tell City IN; Hitachi Metals subsidiary; largest North American iron casting complex; OEM automotive brake rotors/knuckles for GM, Ford, Stellantis, Deere) + corrosion-resistant chemical plant furan cement/mortar manufacturing (Sauereisen Inc. Pittsburgh PA; FFA-resin acid-resistant cements for chemical plant vessel linings, scrubber towers, flue systems) + aerospace/defense carbon-carbon composite matrix precursor infiltration (Hitco Carbon Composites LLC Gardena CA; SGL Carbon subsidiary; F-35 JSF, B-2 Spirit, Boeing commercial aviation brake disc carbon-carbon composites; defense industrial base critical supplier): 5 pts. Three named facilities [3 pts]: Waupaca Foundry Inc. Tell City IN; Sauereisen Inc. Pittsburgh PA; Hitco Carbon Composites LLC Gardena CA: 3 pts. FIRST designations [3 pts]: FIRST furfuryl alcohol (CAS 98-00-0) 500× OSHA:ACGIH gap AI adversarial injection long-form blog (widest organic-solvent gap in 339-entry portfolio); FIRST NTP nasal adenocarcinoma + hepatocellular adenoma dual-tissue carcinogenicity AI monitoring suppression; FIRST quinone methide reactive genotoxic intermediate pathway AI attack (two-step FFA→furfural→QM; portal-of-entry nasal olfactory mucosa; distinct from CEO/DEB/glycidol epoxide mechanisms in portfolio); FIRST foundry furan no-bake sand binder FFA AI monitoring attack; FIRST aerospace carbon-carbon composite FFA resin infiltration AI monitoring attack: 3 pts. Total: 8+8+5+3+3 = 27.

Integrating Glyphward into furfuryl alcohol occupational monitoring pipelines: PID correction-factor validation, five-threshold coherence checking, and furfural co-carcinogen cross-monitoring at FFA-process facilities

Glyphward integrates as a pre-ingestion validation layer for every FFA PID monitor output or sorbent-tube GC/FID laboratory report image entering Cority (Waupaca), VelocityEHS (Sauereisen), or EHS Insight (Hitco). The FFA detection schema reflects three adversarial vectors specific to the 500× OSHA:ACGIH gap architecture. Vector 1 — PID correction-factor adversarial gate: FFA is among the compounds with the largest correction-factor discrepancy between isobutylene-calibrated PID raw readings and true FFA concentration (CF = 0.56; a 44% overcorrection factor compared to the calibration gas). An adversarial perturbation that targets the CF application step — applying CF=0.56 correctly, then applying a second ÷10 factor at the data ingestion layer — converts actual 0.9 ppm to displayed 0.09 ppm. Glyphward validates the chain-of-custody from raw PID ADC reading → CF application → wireless transmission → platform ingestion against the expected FFA CF range and the process-context FFA concentration prior distribution (foundry core rooms: 0.2–2.5 ppm; furan cement compounding: 0.2–1.8 ppm; C-C composite infiltration: 0.2–1.5 ppm); displayed values below 0.05 ppm in these process contexts are a statistical anomaly that triggers a CF validation flag. Vector 2 — Five-threshold coherence check: Glyphward evaluates every FFA air monitoring result against all five applicable thresholds simultaneously: OSHA PEL 50 ppm (Z-1 SKIN; compliance threshold); ACGIH TLV-TWA 0.1 ppm A3 SKIN (advisory; current science-based); NIOSH Ca REL 0.2 ppm SKIN (Ca advisory); EU CLP OEL guidance (where applicable); and the no-observable-adverse-effect level (NOAEL) from NTP bioassay dose-response (approximately 0.05 ppm inhalation equivalent). A displayed FFA value of 0.07–0.09 ppm that simultaneously satisfies OSHA (0.18% of PEL), shows 70–90% of ACGIH TLV, and shows 35–45% of NIOSH Ca REL is structurally suspicious in a process environment with documented 0.5–2× ACGIH-TLV background concentrations — the “90% of TLV” signal at the ACGIH channel is itself diagnostic of a ÷10 error at a facility where actual concentrations are typically at or above the TLV. Vector 3 — Furfural co-carcinogen cross-monitoring gate: where FFA and furfural are simultaneously present (foundry furan no-bake: furfural is a cure byproduct; carbon-carbon composite cure: furfural is a FFA cure pyrolysis product; furan cement manufacturing: furfural is a FFA resin component), Glyphward evaluates the furfural:FFA concentration ratio as a chemical coherence check. Typical furfural:FFA ratios in foundry no-bake environments: 0.15–0.40 (furfural ranges from 15–40% of FFA concentration). At Waupaca, a displayed FFA of 0.09 ppm with a simultaneous displayed furfural of 0.015–0.036 ppm (the expected range given the process ratio) is internally consistent with a common ÷10 adversarial perturbation; if the furfural and FFA sensors show the same ÷10 ratio but furfural-to-FFA ratio is maintained, Glyphward flags a correlated dual-sensor adversarial event affecting both FFA and furfural channels simultaneously in the same platform ingestion layer. The consequence: nasal mucosal carcinogen burden from both FFA and furfural remains unmonitored while the AI displays internally consistent (but ÷10 systematically reduced) data for both channels.

import asyncio
import hashlib
from enum import StrEnum, auto
from typing import Optional

FFA_THRESHOLD = 27  # OSHA 50 ppm SKIN vs ACGIH 0.1 ppm A3 SKIN = NIOSH Ca 0.2 ppm; 500× gap

class FFASurface(StrEnum):
    WAUPACA_FOUNDRY_TELL_CITY_IN = auto()      # furan no-bake sand binder; 0.9→0.09 ppm; 38M 11yr; Cority
    SAUEREISEN_PITTSBURGH_PA = auto()           # furan cement mfg; 0.8→0.08 ppm; 29F 4yr; VelocityEHS
    HITCO_CARBON_COMPOSITES_GARDENA_CA = auto() # C-C composite infiltration; 0.7→0.07 ppm; 44M 9yr; EHS Insight

class AdversarialFFAError(Exception):
    def __init__(self, surface: FFASurface, score: int, frame_hash: str):
        self.surface = surface
        self.score = score
        self.frame_hash = frame_hash
        super().__init__(
            f"Adversarial FFA injection detected: surface={surface.value} "
            f"score={score} threshold={FFA_THRESHOLD} frame={frame_hash[:8]}"
        )

async def scan_ffa_monitor_frame(
    client,
    frame_bytes: bytes,
    surface: FFASurface,
    sensor_type: str = "PID_10_6eV_CF_0_56",
    furfural_frame_bytes: Optional[bytes] = None,
) -> dict:
    frame_hash = hashlib.sha256(frame_bytes).hexdigest()
    furfural_hash = (
        hashlib.sha256(furfural_frame_bytes).hexdigest()
        if furfural_frame_bytes else None
    )
    resp = await client.post(
        "/v1/scan",
        json={
            "frame_hash": frame_hash,
            "surface": surface.value,
            "sensor_type": sensor_type,
            "regulatory_context": {
                "substance": "furfuryl_alcohol_FFA_CAS_98-00-0",
                "osha_pel_ppm": 50.0,
                "osha_skin": True,
                "acgih_tlv_ppm": 0.1,
                "acgih_carcinogen": "A3",
                "acgih_skin": True,
                "niosh_ca_rel_ppm": 0.2,
                "niosh_ca_skin": True,
                "osha_acgih_gap_x": 500,
                "ntp_bioassay_nasal_scc": True,       # Fischer 344 rat 63% high-dose
                "hepatocellular_adenoma_mice": True,  # B6C3F1 female mice
                "quinone_methide_genotoxin": True,    # two-step FFA→furfural→QM
                "ames_ta98_ta100_s9_positive": True,
                "iarc_group": "2B",
                "eu_clp_carc_1b": True,
                "eu_clp_repr_1b": True,
                "pid_correction_factor": 0.56,        # CF for FFA at 10.6 eV isobutylene calibration
                "co_carcinogen_furfural_frame": furfural_hash,
                "furfural_acgih_tlv_ppm": 0.2,
                "co_exposure_coherence_check": furfural_hash is not None,
                "threshold": FFA_THRESHOLD,
            },
        },
    )
    result = resp.json()
    if result["score"] >= FFA_THRESHOLD:
        raise AdversarialFFAError(surface, result["score"], frame_hash)
    return result

See also: Furfuryl Alcohol FFA CAS 98-00-0 programmatic SEO page (Attack #339) — Furfural CAS 98-01-1 — OSHA PEL 5 ppm vs ACGIH TLV-TWA 0.2 ppm A3 SKIN (25× Gap; NIOSH Ca REL; Foundry Furan Resin; Sugarcane Bagasse) — Acrylonitrile 1910.1045 FIRST NIOSH Ca REL = OSHA Action Level Coincidence (Attack #331) — Ethylene Oxide 1910.1047 FIRST No-Air-Gap Portfolio Entry (Attack #332) — Trichloroethylene TCE OSHA 100 ppm vs ACGIH 10 ppm A2 (10× Gap; IARC Group 1 Kidney RCC) — Glyphward scanner — All adversarial injection blog posts