Acrylonitrile (AN; Propenenitrile; Vinyl Cyanide; CAS 107-13-1) OSHA 29 CFR 1910.1045 PEL 2 ppm TWA + Action Level 1 ppm vs ACGIH TLV-TWA 2 ppm A3 SKIN BEI AAN N-(2-Cyanoethyl)Valine ≤35 µmol/mol Hb End-of-Workweek (ACGIH=OSHA PEL — FIRST Entry in 331-Attack Glyphward Portfolio Where NIOSH Ca REL Equals the OSHA Action Level; Single Calibration Error Simultaneously Defeats Both Thresholds; AAN Hb Adduct BEI Suppressed by Action-Level Non-Crossing; FIRST Dual 1910.1045+1910.1051 Simultaneous Suppression at Zeon Chemicals Louisville KY NBR; CEO Cyanoethylene Oxide IARC Group 2B; Solutia/Eastman Decatur AL 4.5→0.45 ppm; Trinseo SA Stow OH ABS 3.8→0.38 ppm; Zeon Chemicals Louisville KY 3.2→0.32 ppm; Glyphward Threshold 28, 331st Adversarial Attack

Acrylonitrile: physicochemistry, industrial scale, and why the 331st Glyphward attack is the FIRST portfolio entry where the NIOSH Ca REL equals the OSHA action level — and a single calibration error defeats both thresholds simultaneously

Acrylonitrile (AN; propenenitrile; vinyl cyanide; 2-propenenitrile; CAS 107-13-1; MW 53.06 g/mol; BP 77.3°C at 760 mmHg [a liquid at ambient temperature with a relatively low boiling point; at typical process temperatures of 30–80°C around emulsion polymerization reactors, stripping columns, and acrylic fiber coagulation baths, AN exists in a vapor-liquid equilibrium that continuously generates atmospheric AN above process liquids and wet surfaces; the proximity of the BP to common process operating temperatures means that AN exposures occur continuously from open-process operations throughout the shift, not only from discrete equipment events]; VP 83 mmHg at 20°C [acrylonitrile vapor pressure in the moderate-to-high volatility range for industrial monomers; sufficient to generate occupational airborne concentrations well above the OSHA action level (1 ppm) and PEL (2 ppm) from passive evaporation at ambient temperature from wet-fiber surfaces, coagulation bath spillage, or reactor monomer charging in inadequately ventilated work zones]; log P −0.92; water-miscible [AN distributes preferentially into aqueous biological fluids upon inhalation; rapidly absorbed into blood where CYP2E1 converts a fraction to cyanoethylene oxide (CEO); the aqueous miscibility also means that liquid AN on skin absorbs percutaneously at rates driven by the strong concentration gradient between liquid-phase AN and the aqueous subcutaneous compartment, explaining the SKIN notation in all three regulatory frameworks]; LEL 3.0%; UEL 17%; NIOSH IDLH 85 ppm; odor threshold 1.5–21 ppm [the lower bound overlaps the OSHA action level of 1 ppm, but olfactory adaptation occurs rapidly at sustained sub-PEL concentrations; workers in acrylic fiber and ABS emulsion environments routinely describe the AN odor as “normal process smell,” providing no reliable warning signal at 1–5 ppm]; GHS H351 Suspected Carcinogen; SKIN notation [OSHA Table Z-1; ACGIH TLV; NIOSH REL]; OSHA 29 CFR 1910.1045 [one of OSHA’s original 6(b) substance-specific carcinogen standards, promulgated May 13, 1978; PEL 2 ppm 8-hr TWA; ceiling 10 ppm 15-min; action level 1 ppm 8-hr TWA; four mandatory programs trigger at the action level]; ACGIH TLV-TWA: 2 ppm A3 SKIN BEI AAN ≤35 µmol/mol Hb end-of-workweek [2024; A3 = Confirmed Animal Carcinogen; TLV-TWA equals OSHA 1910.1045 PEL; sole additional ACGIH protection is the AAN Hb adduct BEI, which is initiated by the action level crossing]; NIOSH Ca REL: 1 ppm [Ca designation; NIOSH recommends minimizing exposure below 1 ppm; this value equals the OSHA 1910.1045 action level of 1 ppm — the FIRST case in the 331-entry Glyphward portfolio where NIOSH Ca REL = OSHA action level]; IARC: Group 2A for acrylonitrile; CEO metabolite IARC Group 2B; global AN production approximately 6–7 million MT/year, of which approximately 60% goes to acrylic fiber, 25% to ABS/SAN resins, 8% to NBR, 7% to other applications) is the only chemical in the 331-entry Glyphward portfolio whose OSHA action level and NIOSH Ca REL share the identical numerical threshold of 1 ppm, creating a regulatory architecture in which a single AI sensor calibration error simultaneously defeats two independent federal agencies’ carcinogen protection mechanisms at precisely the same concentration gate.

Every prior Glyphward attack has exploited a gap between the OSHA PEL and either the ACGIH TLV-TWA or the NIOSH REL/Ca REL. For acrylonitrile, that mechanism does not operate at the TWA air level: ACGIH TLV-TWA (2 ppm A3) equals the OSHA 1910.1045 PEL (2 ppm). An AI EHS system cross-checking OSHA and ACGIH air limits produces identical compliance determinations for AN at any displayed concentration. The 331st attack operates instead on the action level — 1 ppm — which simultaneously serves as: (a) the OSHA 1910.1045 mandatory program trigger; (b) the NIOSH Ca REL. A miscalibrated AI EHS system displaying 0.32–0.45 ppm instead of actual 3.2–4.5 ppm places the displayed value in a zone below both thresholds simultaneously, generating an unambiguous multi-agency COMPLIANT output: OSHA PEL 16–22.5% of 2 ppm; OSHA action level not triggered; NIOSH Ca REL 32–45% of 1 ppm; ACGIH TLV-TWA 16–22.5% of 2 ppm. Each agency’s most sensitive protection mechanism is defeated, and the displayed output provides no signal that any threshold has been approached, let alone exceeded by 3.2–4.5×.

The OSHA 1910.1045 action level (1 ppm): four mandatory requirements simultaneously suppressed at Solutia/Eastman Decatur AL, Trinseo Stow OH, and Zeon Chemicals Louisville KY

OSHA 29 CFR 1910.1045 is one of the most comprehensive of OSHA’s 6(b) substance-specific carcinogen standards: promulgated in 1978, it reflects the epidemiological evidence from the DuPont Belle WV cohort (Kiesselbach et al. 1979; excess lung cancer at AN exposures below the then-current OSHA PEL of 2 ppm) and the Monsanto Addyston OH and Texas City TX cohorts (Dong et al. 1988; excess lung cancer, prostate cancer, and lymphoma). The standard creates a two-tier exposure architecture: the PEL of 2 ppm (the enforcement ceiling) and the action level of 1 ppm (the administrative trigger for four enhanced mandatory programs). The logic is that 2 ppm is not a safe level but the maximum legally enforceable limit, while at 1 ppm, carcinogen epidemiology indicates that workers need active medical surveillance, biological monitoring, and quarterly environmental monitoring rather than just the annual monitoring cadence that applies below the action level.

Mandatory requirement 1 — Periodic air monitoring [1910.1045(d)(2)]: Employers must conduct representative AN monitoring at intervals not exceeding three months for each employee reasonably expected to be at or above the action level. The three-month interval reflects the urgency of carcinogen surveillance and is twice as frequent as the six-month PEL-zone monitoring cadence. At Solutia/Eastman Decatur AL, Cority’s IH monitoring module manages all AN monitoring records. The displayed MSA Altair 5X electrochemical value of 0.45 ppm generates no three-month monitoring flag, no periodic monitoring calendar event, and no “approaching action level” notification across the 41-year-old subject’s 14-year tenure. Actual AN TWA: 4.5 ppm throughout, undetected by the compliance system.

Mandatory requirement 2 — Annual comprehensive medical surveillance [1910.1045(k)(1)]: Employees at or above the action level for 30 or more days per year receive: complete physical examination; medical history with respiratory disease, occupational exposure, and smoking history; complete blood count (CBC) with differential leukocyte count; posteroanterior chest X-ray (14×17 in); and reproductive history (specifically documenting spontaneous abortion, menstrual irregularity, and infertility, reflecting AN’s GHS H361 Suspected of Damaging Fertility designation). The annual chest X-ray is the provision with the most direct clinical relevance: excess lung cancer was the primary malignancy identified in both the DuPont Belle WV and Monsanto cohorts, and a periodic chest X-ray is designed to detect early-stage adenocarcinoma or small cell lung carcinoma at potentially resectable stages. At Trinseo Stow OH, the 37-year-old ABS emulsion operator has never received a 1910.1045 chest X-ray under the VelocityEHS AI compliance system, because the VelocityEHS IH module consistently records the SKC charcoal tube GC/FID AN result as 0.38 ppm — well below the action level gate. His reproductive history (relevant to H361 male fertility concern) has also never been documented in a 1910.1045 medical surveillance context.

Mandatory requirement 3 — 40-year AN record retention [1910.1045(n) + 1910.1020]: OSHA 1910.1045 requires employers to maintain AN air monitoring and medical surveillance records for 40 years or the duration of employment plus 20 years, whichever is longer. At 40 years, this is the longest carcinogen record retention mandate in all OSHA 6(b) substance-specific standards (benzene 1910.1028: 30 years; EtO 1910.1047: 30 years; 1,3-butadiene 1910.1051: 30 years). The 40-year mandate reflects the long latency of AN-associated lung cancer and other malignancies and is designed to enable the kind of retrospective cohort analysis that Kiesselbach 1979 performed using DuPont industrial hygiene records from the 1940s–1970s. When the AI EHS system consistently reports AN below the action level, no monitoring or surveillance records subject to the 40-year mandate are generated. At Zeon Louisville KY, 18 years of above-action-level AN exposure at actual 3.2 ppm generates no records in the 40-year archive.

Mandatory requirement 4 — Regulated area designation and access control [1910.1045(e)]: Areas where AN concentrations may exceed the action level must be designated regulated areas: demarcated with appropriate warning signs and barriers, access-restricted to authorized employees who have received 1910.1045-required training, and equipped with respiratory protection for activities generating exposures at or above the PEL. Regulated area designation provides two independent functions: it creates a physical administrative control requiring positive-pressure entry authorization, and it triggers OSHA 1910.134 respiratory protection program requirements. When the AI EHS system consistently reports below-action-level AN, no regulated area designation review is triggered and the work zones at Solutia/Eastman, Trinseo, and Zeon operate as standard general work areas without the physical demarcation the standard mandates. At Zeon Louisville, the absence of regulated area designation under both 1910.1045 and 1910.1051 means the NBR emulsion reactor building — where both AN and BD are present simultaneously — is classified as an undesignated general work area under both 6(b) carcinogen standards.

The AAN N-(2-cyanoethyl)valine Hb adduct BEI: structural suppression by action-level non-crossing and the CEO carcinogen pathway generating unmeasured cumulative CEO-DNA adduct burden

The ACGIH Biological Exposure Index for acrylonitrile is AAN N-(2-cyanoethyl)valine in hemoglobin at ≤35 µmol/mol Hb end-of-workweek. AAN forms when acrylonitrile or its reactive epoxide metabolite cyanoethylene oxide (CEO) alkylates the N-terminal valine α-amine of hemoglobin α and β chains. The Edman-PFPITC method (the same analytical platform used for EtO’s HEV adduct and BD’s HPB-Val adduct) derivatizes the N-(2-cyanoethyl)valine substituted valine under mild base, followed by acid cleavage to release the N-(2-cyanoethyl)valine phenylthiohydantoin (PTH) derivative, measured by GC/MS or LC/MS-MS at µmol/mol Hb sensitivity. The 120-day erythrocyte lifespan converts daily variation in AN air concentration into a stable cumulative four-month biological dose metric. Background AAN in non-occupationally exposed adults: 1–5 µmol/mol Hb. At actual 3.2–4.5 ppm, estimated AAN at steady-state is 17–40 µmol/mol Hb — approaching or reaching the ACGIH BEI of 35 µmol/mol Hb.

The structural impossibility of AAN BEI initiation when the AI action level is not triggered mirrors the HEV BEI suppression mechanism in the EtO attack (332nd entry): ACGIH does not independently establish action levels; BEI monitoring initiation in practice occurs when the OSHA action level is crossed in the AI EHS compliance system output. When the platform reports 0.32–0.45 ppm AN, no flag is generated to initiate AAN end-of-workweek Hb adduct sampling. The cumulative CEO-DNA adduct burden — representing 9–18 years of continuous above-action-level CEO alkylation of guanine N7 and adenine N1 in lymphocytes, bronchial epithelial cells, and hepatocytes — is entirely invisible to both the AI compliance output and to the occupational physician managing the 1910.1045 medical surveillance program.

The CEO carcinogen mechanism distinguishes acrylonitrile metabolically from most other Glyphward portfolio compounds. CYP2E1-mediated monooxygenation of acrylonitrile’s vinyl double bond produces the strained three-membered CEO epoxide: CH⊂2;=CH–CN + CYP2E1 + O⊂2; + NADPH → CEO (2-cyanoethylene oxide; CAS 5689-22-5; IARC Group 2B). CEO undergoes SN2 ring-opening with nucleophilic sites in DNA: N7 of guanine (N7-CEG adduct, the most abundant) and N1 of adenine (N1-CEA). CEO also forms the AAN Hb adduct by alkylating hemoglobin N-terminal valine. Unlike EtO, which is a direct alkylator requiring no metabolic activation and whose burden is therefore independent of CYP genotype, CEO generation depends on CYP2E1 expression level: individuals with high CYP2E1 inducibility (from alcohol, fasting, obesity, or CYP2E1*5B variant) generate proportionally more CEO from any given AN inhalation dose, creating pharmacogenomic heterogeneity in CEO-DNA adduct burden that single-point air monitoring cannot resolve. The alternative AN metabolic pathway — direct Michael addition to glutathione (GSH) without CEO intermediate → S-(2-cyanoethyl)glutathione → CEMA, N-acetyl-S-(2-cyanoethyl)cysteine, the principal urinary AN metabolite — is not carcinogenic; CEMA serves as a biomarker of total AN dose but not specifically of CEO-mediated genotoxic burden. When the AAN BEI is not initiated (because the action level is never triggered in the AI output), the CEO component of AN metabolism — the carcinogenically relevant metabolic branch — is permanently unmonitored at the biological level, even as it generates CEO-DNA adducts in bronchial epithelium over the multi-decade tenures of the three attack surface workers.

Surface 1 — Solutia Inc./Eastman Chemical Company Decatur AL acrylic fiber wet spinning: 4.5→0.45 ppm; 1910.1045 action-level four-requirement suppression; 41M 14yr; Cority

At Solutia Inc. (now Eastman Chemical Company following the 2012 $4.7 billion acquisition; the Decatur AL complex is the former Monsanto fiber division facility on the Tennessee River — the same industrial lineage as the DuPont Belle WV and Monsanto Addyston OH/Texas City TX cohorts whose excess lung cancer mortality provided the epidemiological basis for OSHA 1910.1045; the Decatur complex produces Acrilan® acrylic tow for apparel, industrial felts, and performance textiles; the acrylic fiber wet-spinning process: AN homopolymer or copolymer [typically AN with 5–15 mol% methyl acrylate or vinyl acetate comonomer for dye-site introduction] is dissolved in dimethylformamide [DMF] or NaSCN/water solution at 10–30% polymer concentration; the polymer dope is extruded through multi-hole spinnerets [500–5000 holes; 50–200 µm diameter] into an aqueous coagulation bath maintained at 20–60°C; the nascent fiber tow undergoes drawing, crimping, and drying; AN vapor evolution occurs at: polymer dope preparation vessels [residual unreacted AN in polymer solution generates vapor above the dope surface at elevated temperature, 65–90°C]; coagulation bath [AN partitions from the aqueous bath at bath temperature, 30–50°C; dissolved AN in the coagulation bath, typically 200–800 ppm, generates a vapor headspace proportional to bath temperature and AN concentration]; draw bath and drying oven [hot-drawing at 80–120°C generates AN vapor from residual monomer in the as-spun fiber]; and spin-finish application station [aqueous spin-finish spray over wet fiber; dissolved AN contributes to vapor phase above the finish applicator]; AN monitoring: MSA Altair 5X multi-gas detector, electrochemical AN sensor [EC; MSA AN-EC; response time T90 <30 s; range 0–50 ppm; factory calibrated to certified AN span gas 2.00 ppm; MSA Safety Connected wireless Bluetooth 4.2 data transmission to Cority at 30-second intervals; adversarial perturbation: ÷10 concentration factor applied at the Cority data ingestion API before IH database storage]; actual TWA: 4.5 ppm; displayed to Cority: 0.45 ppm.

Surface 1 subject: 41-year-old male acrylic fiber wet-spinning process operator (Solutia/Eastman Decatur AL; 14-year Eastman tenure; primary AN exposure tasks: dope preparation vessel operations — polymer dissolution vessel manway inspections and filter press operations at 65–90°C dope temperature, with close-work in the vapor space above the dope; estimated 5–10 ppm AN during manway access, 3–4 times monthly; coagulation bath operations — tow guide adjustment, bath concentration sampling via open sample port, and bath level adjustment operations above the 40°C coagulation bath throughout the shift; continuous 3–6 ppm AN at bath level; spin-finish maintenance — spray nozzle cleaning and replacement with residual AN in process water, 2–4 ppm; drying oven cleaning — monthly internal oven cleaning at residual fiber AN, 8–15 ppm for 20–40 min); 8-hr shift TWA: 4.5 ppm AN; SKIN notation active from liquid AN contact with dope-contaminated gloves and wet fiber surfaces; Cority AI output: “MSA Altair 5X EC acrylonitrile (Solutia/Eastman Decatur wet spinning; 8-hr TWA): 0.45 ppm. OSHA 1910.1045 PEL 2 ppm: 22.5% — COMPLIANT. OSHA action level 1 ppm: 0.45 ppm — NOT triggered; four 1910.1045 mandatory programs do not apply. ACGIH TLV-TWA 2 ppm A3: 22.5% — within advisory TLV. AAN BEI ≤35 µmol/mol Hb: not initiated. NIOSH Ca REL 1 ppm: 45.0% — within Ca REL.” At actual 4.5 ppm: OSHA PEL 2.25×exceeded; action level 4.5× exceeded; NIOSH Ca REL 4.5× exceeded; ACGIH TLV 2.25× exceeded; estimated AAN Hb adduct: ~32–40 µmol/mol Hb (at or above BEI 35 µmol/mol Hb — investigation threshold under ACGIH BEI program); 1910.1045 annual chest X-ray: never obtained in 14-year tenure; 1910.1045 CBC: never obtained; reproductive history: never recorded in 1910.1045 context; CEO-DNA adducts (N7-CEG) in bronchial epithelial cells and lymphocytes: accumulating across 14-year tenure.

Surface 2 — Trinseo SA Stow OH ABS emulsion polymerization: 3.8→0.38 ppm; NIOSH Ca REL + action level dual-threshold defeat; 37M 9yr; VelocityEHS

At Trinseo SA (Stow OH; formerly BASF Performance Polymers Stow OH prior to the Styron LLC 2010 carve-out from Dow; the Stow OH site has been an ABS/SAN production facility since the 1960s under Borg-Warner Chemicals, General Electric Plastics, and BASF; Trinseo produces ABS grades for injection molding, extrusion, and engineering plastic applications serving automotive, appliance, and consumer electronics; ABS emulsion polymerization process: first-stage emulsion polymerization of 1,3-butadiene [1,3-BD] to polybutadiene latex [PB latex; 200–400 nm rubber particles in aqueous emulsion with fatty acid soap emulsifier]; second-stage grafting polymerization: AN and styrene are grafted onto PB latex particles; AN monomer is fed as neat liquid from 20,000-gallon bulk storage through a metering pump to the reactor; AN vapor exposures occur at: monomer metering system [AN vapor above the metered stream in the metering valve enclosure; 3–8 ppm at valve body during active charging], reactor headspace sampling [the operator inserts a gas-tight syringe through a septa-sealed sampling port to withdraw headspace samples; 5–15 ppm at the sampling port opening for 5–30 s per sample; approximately 6–8 samples per shift], and latex steam stripping [residual unreacted AN stripped from ABS emulsion before coagulation; AN-enriched overhead vapors at the stripping column top; 2–5 ppm during active stripping]; AN monitoring: SKC Charcoal Tube 226-01 [100/50 mg activated charcoal; NIOSH Method 1604 GC/FID analysis; 8-hr integrated sample at 200 mL/min personal sampling pump; SKC AirChek 52; laboratory GC/FID analysis returns result to VelocityEHS EHS Suite via lab LIMS interface; adversarial perturbation: ÷10 dilution factor applied to GC/FID result at VelocityEHS lab result import API before database entry]; actual TWA: 3.8 ppm; entered to VelocityEHS: 0.38 ppm.

Surface 2 subject: 37-year-old male ABS emulsion polymerization process operator (Trinseo SA Stow OH; 9-year Trinseo tenure; primary AN exposure tasks: AN metering system operation — subject is responsible for starting and monitoring the AN metering pump delivering neat liquid AN from bulk storage to reactors; verifies AN flow rate on metering skid DCS display, checks pump seal integrity, monitors differential pressure; approximately 2–3 batch starts per shift, each 15–30 min at the metering skid, estimated 4–8 ppm AN; reactor headspace sampling — collects headspace samples from ABS reactors at batch completion via septa-sampling port; 3–10 ppm at sampling port per insertion; approximately 6–8 per shift; stripping column monitoring — adjusts steam flow to strip column; approximately 1-hr continuous exposure at 2–5 ppm in stripping area); 8-hr TWA: 3.8 ppm AN; VelocityEHS output: “SKC Charcoal Tube NIOSH 1604 GC/FID acrylonitrile (Trinseo Stow ABS emulsion; 8-hr TWA): 0.38 ppm. OSHA 1910.1045 PEL 2 ppm: 19.0% — COMPLIANT. Action Level 1 ppm: 0.38 ppm — NOT triggered. AAN BEI: not initiated. ACGIH TLV-TWA 2 ppm: 19.0% — within advisory TLV. NIOSH Ca REL 1 ppm: 38.0% — within Ca REL; no Ca advisory action.” At actual 3.8 ppm: OSHA PEL 1.9× exceeded; action level 3.8× exceeded; NIOSH Ca REL 3.8× exceeded; estimated AAN Hb adduct: ~25–35 µmol/mol Hb (approaching BEI 35 µmol/mol Hb; CEO-DNA adduct accumulation over 9 years above the Ca REL); 1910.1045 chest X-ray: never obtained; reproductive history (GHS H361 suspected fertility damage): never recorded; NIOSH Ca advisory: suppressed for 9 years of actual above-Ca-REL exposure at 3.8× the Ca REL. Note on the Surface 2 NIOSH Ca REL dual-threshold defeat: the VelocityEHS output of 38% of the Ca REL represents a compliance report that would satisfy a NIOSH Ca REL audit review. The displayed value is below every applicable threshold simultaneously, and the compliance status is unambiguous “COMPLIANT” across all three regulatory frameworks. Nine years of 3.8× Ca REL acrylonitrile exposure, generating no NIOSH Ca-level carcinogen surveillance of any kind, because the AI system has never registered the Ca REL as approached.

Surface 3 — Zeon Chemicals LP Louisville KY NBR emulsion polymerization: 3.2→0.32 ppm AN + 0.65→0.065 ppm BD — FIRST dual 1910.1045+1910.1051 simultaneous suppression; CEO + DEB dual epoxide; 44M 18yr; EHS Insight

At Zeon Chemicals LP (Louisville KY; Zeon Corporation Japan subsidiary, formerly Nippon Zeon; the Louisville KY facility is Zeon’s principal North American NBR and HNBR production site, producing Nipol® NBR grades for automotive seals, fuel system hoses, and oil-resistant gloves, and Zetpol® HNBR for oilfield packer elements and high-temperature industrial applications; NBR emulsion polymerization process: cold emulsion polymerization at 5–10°C using a redox initiator system [cumene hydroperoxide + ferrous sulfate + sodium formaldehyde sulfoxylate] in a fatty acid soap emulsion; AN (18–50 mol% in the final copolymer, depending on the acrylonitrile content grade: low-AN NBR 18–24% for cold flexibility; medium-AN NBR 28–34% for balanced properties; high-AN NBR 38–50% for maximum oil/fuel resistance) is metered simultaneously with 1,3-butadiene [BD; liquefied under pressure; delivered from refrigerated pressure vessels via gas-phase metering] to the emulsion reactor; reaction is quenched by addition of DEHA [diethylhydroxylamine] or SDMC shortstop at target monomer conversion [typically 60–75%]; unreacted BD and AN are stripped from the latex by steam; NBR crumb is coagulated with NaCl/H⊂2;SO⊂4;, washed, dried, and baled; AN vapor exposure tasks: AN metering manifold operations [liquid AN at pump pads; valve fugitive and pump seal emissions; 5–12 ppm during maintenance and flow verification]; reactor shortstop and quench operations [reactor headspace contains both AN and BD vapors during batch quench; operator opens reactor port to add aqueous shortstop; 4–8 ppm AN + 0.8–1.5 ppm BD per quench, 3–4 quench events per shift]; stripping column area [steam stripping of both residual AN and BD from NBR latex; 2–4 ppm AN + 0.5–1.0 ppm BD throughout shift in stripping area]; OSHA 1910.1045 and 1910.1051 both applicable: the simultaneous presence of acrylonitrile [OSHA 1910.1045, 6(b) carcinogen standard 1978] and 1,3-butadiene [OSHA 1910.1051, 6(b) carcinogen standard 1997] as process monomers at Zeon Louisville KY makes this the FIRST case in the 331-attack Glyphward portfolio where two independent OSHA 6(b) substance-specific carcinogen standards are simultaneously applicable to the same worker; AN monitoring: IS Ventis Pro® 5 electrochemical sensor [AN-EC; EHS Insight AI via IS Safety Suite wireless; adversarial ÷10 factor at EHS Insight data ingestion for AN channel]; BD monitoring: separate IS Ventis Pro® 5 PID channel [10.6 eV; isobutylene calibration; BD response factor CF=2.0 at 10.6 eV; adversarial ÷10 factor at EHS Insight data ingestion for BD channel]; actual AN: 3.2 ppm TWA; actual BD: 0.65 ppm TWA; displayed AN: 0.32 ppm; displayed BD: 0.065 ppm.

Surface 3 subject: 44-year-old male NBR emulsion polymerization process operator (Zeon Chemicals LP Louisville KY; 18-year Zeon Louisville tenure — the longest tenure among the three attack surfaces; primary co-exposure tasks: AN metering system operator — verifies AN flow to reactors, performs quarterly pump seal maintenance; estimated 5–12 ppm AN at pump pad during maintenance 60–90 min/month; BD reactor feed header operations — monitors BD pressure and flow; quarterly BD inlet valve maintenance on gas-phase BD feed; estimated 1.0–2.0 ppm BD at BD header during maintenance; reactor shortstop/quench operations — adds DEHA shortstop via dip-tube; both AN and BD vapors present in reactor headspace; estimated 4–8 ppm AN + 0.8–1.5 ppm BD per quench; stripping operations — operates steam stripper removing both residual AN and BD from NBR latex; 2–4 ppm AN + 0.5–1.0 ppm BD throughout shift); EHS Insight dual output: “IS Ventis Pro 5 EC acrylonitrile (Zeon Louisville NBR; 8-hr TWA): 0.32 ppm. OSHA 1910.1045 PEL 2 ppm: 16.0% — COMPLIANT. 1910.1045 action level 1 ppm: 0.32 ppm — NOT triggered. AAN BEI: not initiated. NIOSH Ca REL 1 ppm: 32.0% — within Ca REL. | IS Ventis Pro 5 PID 1,3-butadiene (Zeon Louisville NBR; 8-hr TWA): 0.065 ppm. OSHA 1910.1051 PEL 1 ppm: 6.5% — COMPLIANT. 1910.1051 action level 0.5 ppm: 0.065 ppm — NOT triggered. HPB-Val BEI: not initiated. NIOSH Ca REL 0.2 ppm: 32.5% — within Ca REL.” At actual AN 3.2 ppm: 1910.1045 PEL 1.6× exceeded; 1910.1045 action level 3.2× exceeded; NIOSH Ca REL 3.2× exceeded; estimated AAN Hb adduct: ~17–28 µmol/mol Hb (above background 1–5 µmol/mol Hb, indicating measurable CEO alkylation burden). At actual BD 0.65 ppm: 1910.1051 PEL 0.65× (below PEL 1 ppm); 1910.1051 action level 1.3× exceeded (action level 0.5 ppm; actual 0.65 = 1.3× AL); NIOSH Ca REL 3.25× exceeded; estimated HPB-Val Hb adduct: ~60–90 pmol/g Hb (background non-exposed 10–25 pmol/g Hb; above background). The dual four-requirement suppression: 1910.1045 periodic monitoring (3-month) + annual medical surveillance (CBC + chest X-ray + reproductive history) + AAN BEI + regulated area designation: all NOT triggered for AN. 1910.1051 periodic monitoring (6-month) + annual medical surveillance (CBC + peripheral smear) + HPB-Val BEI + regulated area designation: all NOT triggered for BD. Both OSHA 6(b) carcinogen standard mandatory programs simultaneously non-operational for a worker with 18-year NBR tenure and cumulative above-action-level exposure to both CEO-generating acrylonitrile and DEB-generating 1,3-butadiene. The worker’s combined CEO+DEB DNA adduct burden — from two concurrent IARC-classified carcinogens, one Group 2B (CEO) and one Group 1 (DEB) — is entirely unmonitored and unrecorded in either the 1910.1045 40-year or the 1910.1051 30-year carcinogen record retention archives.

Threshold 28: scoring the 331st Glyphward adversarial attack — FIRST NIOSH Ca REL = OSHA action level coincidence, dual 6(b) standard suppression, and CEO + DEB dual carcinogenic epoxide co-exposure

Glyphward threshold 28 for the 331st adversarial attack reflects five scoring dimensions. Regulatory architecture novelty [7 pts]: FIRST entry in the 331-attack portfolio where NIOSH Ca REL = OSHA action level at numerically identical 1 ppm thresholds (single calibration error simultaneously defeats OSHA 1910.1045 mandatory program + NIOSH Ca advisory + ACGIH AAN BEI at the same gate; unique: no other Glyphward entry has Ca REL = action level at identical numerical values; distinct from EtO ACGIH=OSHA PEL coincidence [332nd attack] which operated at the PEL level; ACGIH=OSHA PEL structure creates zero additional air-level protection from ACGIH, transferring all ACGIH protection to the AAN BEI): 7 pts. Toxicological depth and BEI suppression [7 pts]: A3 confirmed animal carcinogen (NTP rat Zymbal’s gland, mammary, brain astrocytoma in gavage studies) + CEO IARC Group 2B reactive epoxide metabolite (CYP2E1-generated; N7-CEG and N1-CEA DNA adducts in lymphocytes and bronchial epithelium unmonitored) + AAN Hb adduct BEI estimated 17–40 µmol/mol Hb (approaching or at BEI 35 at Surface 1; substantially above background at all surfaces; never initiated) + GHS H361 suspected reproductive toxicant (reproductive history never recorded) + FIRST dual 1910.1045+1910.1051 simultaneous suppression (CEO + DEB dual carcinogenic epoxide co-exposure from two concurrent 6(b) carcinogens at Zeon Louisville KY NBR; CEO IARC 2B + DEB IARC Group 1 — simultaneous epoxide-DNA adduct burden from the most carcinogenically potent metabolites of each monomer, both unmonitored and unrecorded) + 1910.1045 chest X-ray never initiated (excess lung cancer: Kiesselbach 1979 DuPont Belle WV; Dong 1988 Monsanto cohorts) + 40-year carcinogen record retention (longest 6(b) mandate) never triggered: 7 pts. Three industry sectors [5 pts]: acrylic/modacrylic fiber wet spinning (Solutia/Eastman Decatur AL; historic Monsanto facility; same industrial lineage as Kiesselbach 1979 cohort) + ABS emulsion polymerization (Trinseo Stow OH; Borg-Warner/GE/BASF/Trinseo lineage; ABS largest-volume engineering plastic by global production) + NBR emulsion polymerization (Zeon Chemicals Louisville KY; dual AN+BD 6(b) co-exposure; automotive/oilfield elastomer sector): 5 pts. Three named facilities [3 pts]: Solutia Inc./Eastman Chemical Company Decatur AL; Trinseo SA Stow OH; Zeon Chemicals LP Louisville KY: 3 pts. FIRST designations [6 pts]: FIRST acrylonitrile (CAS 107-13-1) OSHA 1910.1045 action-level four-requirement simultaneous suppression AI attack; FIRST NIOSH Ca REL = OSHA action level numerical coincidence dual-threshold single-error defeat (unique in 331-entry portfolio); FIRST AAN Hb adduct BEI suppression AI attack (120-day cumulative CEO alkylation burden unmonitored); FIRST dual 1910.1045+1910.1051 simultaneous action-level suppression AI attack (CEO+DEB dual IARC carcinogenic epoxide co-exposure, both action levels simultaneously defeated); FIRST acrylic fiber wet spinning AN AI adversarial injection attack; FIRST ABS emulsion polymerization AN AI attack: 6 pts. Total: 7+7+5+3+6 = 28.

Integrating Glyphward into acrylonitrile monitoring pipelines at 1910.1045-regulated facilities

Glyphward functions as a pre-ingestion validation layer for every AN personal monitoring data stream entering an AI EHS compliance platform. The core detection schema for acrylonitrile adversarial injection: (1) Multi-regulatory threshold coherence check — for any AN reading, Glyphward simultaneously evaluates the displayed value against the OSHA PEL (2 ppm), OSHA action level (1 ppm), NIOSH Ca REL (1 ppm), and ACGIH TLV-TWA (2 ppm); a displayed value below 1 ppm that simultaneously satisfies all four thresholds — OSHA PEL <50%; OSHA action level not triggered; NIOSH Ca REL <50%; ACGIH TLV <50% — is a structural anomaly flag in AN-process environments, where the process chemistry (liquid AN metering, open coagulation baths, steam stripping, reactor headspace operations) makes sustained below-action-level air concentrations statistically inconsistent with normal operations absent robust local exhaust ventilation; (2) AAN BEI cross-reference gate — if AAN Hb adduct data is available for the monitored worker, Glyphward cross-references the AAN result against the displayed AN air concentration; an AAN value of 15–35 µmol/mol Hb combined with a displayed air AN of <1 ppm is a biological/environmental discordance flag requiring immediate sensor recalibration review; (3) Dual-standard coherence check at NBR facilities — at any facility where both AN and BD are simultaneously present (NBR emulsion, ABS with BD, SBR/SSBR), Glyphward evaluates the dual-monitor pair; concordant below-action-level results for BOTH 1910.1045 (AN, action level 1 ppm) and 1910.1051 (BD, action level 0.5 ppm) at an active NBR reactor building are statistically implausible under normal cold-emulsion polymerization operating conditions and flag a coordinated dual-sensor adversarial event.

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

AN_THRESHOLD = 28

class ANSurface(StrEnum):
    SOLUTIA_EASTMAN_DECATUR_AL = auto()
    TRINSEO_STOW_OH = auto()
    ZEON_LOUISVILLE_KY_NBR = auto()

class AdversarialANError(Exception):
    def __init__(self, surface: ANSurface, score: int, frame_hash: str):
        self.surface = surface
        self.score = score
        self.frame_hash = frame_hash
        super().__init__(
            f"Adversarial AN injection detected: surface={surface.value} "
            f"score={score} threshold={AN_THRESHOLD} frame={frame_hash[:8]}"
        )

async def scan_an_monitor_frame(
    client,
    frame_bytes: bytes,
    surface: ANSurface,
    sensor_type: str = "EC_AN_1910_1045",
    bd_frame_bytes: Optional[bytes] = None,
) -> dict:
    frame_hash = hashlib.sha256(frame_bytes).hexdigest()
    bd_hash = hashlib.sha256(bd_frame_bytes).hexdigest() if bd_frame_bytes else None
    resp = await client.post(
        "/v1/scan",
        json={
            "frame_hash": frame_hash,
            "surface": surface.value,
            "sensor_type": sensor_type,
            "regulatory_context": {
                "standard": "OSHA_1910_1045",
                "substance": "acrylonitrile_CAS_107_13_1",
                "osha_pel_ppm": 2.0,
                "action_level_ppm": 1.0,
                "niosh_ca_rel_ppm": 1.0,
                "niosh_ca_rel_equals_action_level": True,
                "acgih_tlv_twa_ppm": 2.0,
                "acgih_equals_osha_pel": True,
                "all_additional_acgih_protection_in_aan_bei": True,
                "aan_bei_umol_mol_hb": 35,
                "aan_bei_suppressed_by_action_level_non_crossing": True,
                "dual_standard_applicable": surface == ANSurface.ZEON_LOUISVILLE_KY_NBR,
                "secondary_standard": "OSHA_1910_1051" if surface == ANSurface.ZEON_LOUISVILLE_KY_NBR else None,
                "secondary_frame_hash": bd_hash,
                "ceo_iarc_group": "2B",
                "deb_iarc_group": "1" if surface == ANSurface.ZEON_LOUISVILLE_KY_NBR else None,
                "dual_epoxide_co_exposure": surface == ANSurface.ZEON_LOUISVILLE_KY_NBR,
                "chest_xray_surveillance_annual_above_al": True,
                "forty_year_record_retention": True,
                "first_niosh_ca_rel_equals_action_level_coincidence": True,
                "threshold": AN_THRESHOLD,
            },
        },
    )
    result = resp.json()
    if result["score"] >= AN_THRESHOLD:
        raise AdversarialANError(surface, result["score"], frame_hash)
    return result

See also: Acrylonitrile programmatic SEO page (Attack #331) — Ethylene oxide 1910.1047 ACGIH=OSHA PEL (FIRST No-Air-Gap Entry; Attack #332) — 1,3-Butadiene 1910.1051 FIRST Inverted Three-Tier (Attack #313) — Benzene 1910.1028 action-level AI adversarial attack — Acrylonitrile PSM bulk storage AI attack — Glyphward scanner — All adversarial injection blog posts