Ethylene Oxide (EtO; Oxirane; CAS 75-21-8) OSHA 29 CFR 1910.1047 PEL 1 ppm TWA + Action Level 0.5 ppm vs ACGIH TLV-TWA 1 ppm A2 BEI HEV N-(2-Hydroxyethyl)Valine ≤35 pmol/g Hb End-of-Workweek (ACGIH=OSHA — FIRST Entry in 332-Attack Glyphward Portfolio Where ACGIH Provides No Additional Inhalation Protection; All Protection Transferred to HEV Hb Adduct BEI; BEI Suppressed by Action-Level Non-Crossing) vs NIOSH Ca REL 0.1 ppm (10× Below OSHA; IARC Group 1 Breast Cancer / NHL / Leukemia; Steenland 2004 NIOSH Mortality Cohort 18,235 Workers; Direct Alkylating Agent; O6-HEG Mutagenic GC→AT; Ascension Saint Francis Hospital Evanston IL CSSD 0.85→0.085 ppm; Sterigenics Willowbrook IL 0.72→0.072 ppm; LyondellBasell La Porte TX EtO Production 0.90→0.09 ppm; Glyphward Threshold 30, 332nd Adversarial Attack

Ethylene oxide: physicochemistry, industrial scale, and why the 332nd Glyphward attack is the FIRST portfolio entry where ACGIH TLV-TWA equals the OSHA PEL — and where all additional protection is embedded in the HEV Hb adduct BEI suppressed by action-level miscalibration

Ethylene oxide (EtO; oxirane; 1,2-epoxyethane; CAS 75-21-8; MW 44.05 g/mol; BP 10.7°C at 760 mmHg [EtO boils just below ambient temperature and exists as a gas at any worksite above 10.7°C — meaning that any EtO container breach, valve seat failure, or sterilizer door seal degradation results in immediate, continuous vapor generation without the rate-limiting evaporation step that governs liquid solvent exposures; the gas is already present at the release event onset]; VP 1,095 mmHg at 20°C [nearly 1.5 atmospheres vapor pressure at ambient temperature — among the highest VP of any regulated occupational carcinogen in the Glyphward portfolio; for comparison, benzene VP is 95 mmHg, trichloroethylene VP is 77 mmHg, formaldehyde VP is 3,886 mmHg as gas; EtO’s extreme volatility means that even trace liquid releases produce immediate high-concentration vapor events]; water solubility miscible [EtO dissolves completely in water at any proportion — this property is responsible for its penetrating sterilization efficacy through wrapped and packaged medical device barriers, but also means that after inhalation, systemic distribution of dissolved EtO into plasma, lymph, and tissue water compartments proceeds essentially instantaneously, delivering the direct alkylating species to target tissues including breast epithelium, bone marrow, and lymphoid progenitor populations without a lag phase]; SKIN notation [percutaneous absorption; aqueous solutions of EtO and gas-phase EtO contact with skin contribute non-negligible systemic dose fractions at occupational concentrations, though inhalation remains the dominant exposure pathway in properly gloved workers]; log P −0.30 [hydrophilic; distributes preferentially into aqueous compartments including blood; this hydrophilicity underlies the high blood:air partition coefficient of EtO (approximately 730 at 37°C), which means that inhaled EtO is efficiently absorbed into the pulmonary circulation and delivered to hemoglobin in circulating erythrocytes, where HEV adduct formation occurs — the very mechanism that makes the HEV biomarker a quantitative surrogate for cumulative systemic EtO dose]; odor threshold 260–700 ppm [the critical safety limitation of all EtO monitoring: EtO’s odor threshold is far above every occupational exposure limit by two to three orders of magnitude; a worker cannot detect EtO by smell at 1 ppm (OSHA 1910.1047 PEL), 0.5 ppm (OSHA action level), or 0.1 ppm (NIOSH Ca REL); there is no olfactory early-warning signal at any concentration that occupational hygiene regulation considers protective; this creates an absolute instrument dependency for EtO monitoring that has no parallel among common industrial solvent vapors — and transforms any calibration error in those instruments into an undetectable, self-reinforcing monitoring failure]; NIOSH IDLH 800 ppm; GHS H350 May Cause Cancer [IARC Group 1; EU Cat. 1A]; H220 Extremely Flammable Gas [GHS Flammable Gas Category 1A; LEL 3.0%; UEL 100% — EtO has a fully flammable range from 3% to 100%; uniquely among common gases, pure EtO vapor can sustain a vapor-phase decomposition explosion in the complete absence of air above approximately 3%; the 100% UEL means there is no upper concentration at which EtO becomes “too rich to ignite” — which is why EtO sterilization facilities are designed with explosion-proof electrical systems, inerting gas blankets over product storage vessels, and dedicated ventilation; but these large-release explosion controls are facility-level engineering measures entirely separate from the personal occupational air monitoring under 1910.1047 that addresses chronic carcinogen inhalation at sub-IDLH concentrations]; OSHA 29 CFR 1910.1047 [one of OSHA’s 6(b) substance-specific carcinogen standards; promulgated 1984; regulates EtO occupational exposures in any establishment in which EtO is used or produced; establishes PEL 1 ppm 8-hr TWA, STEL (excursion limit) 5 ppm 15-min, and action level 0.5 ppm 8-hr TWA with associated four-requirement mandatory program cascade]; ACGIH TLV-TWA: 1 ppm A2 BEI HEV ≤35 pmol/g Hb end-of-workweek [A2 = Suspected Human Carcinogen; the ACGIH TLV-TWA of 1 ppm is identical to the OSHA 1910.1047 PEL of 1 ppm — the FIRST and only entry in the 332-attack Glyphward portfolio where ACGIH provides no additional inhalation protection at the air concentration level beyond the applicable OSHA standard; the entirety of ACGIH’s additional protection comes from the HEV Hb adduct BEI]; NIOSH Ca REL: 0.1 ppm 10-hr TWA [Ca designation — NIOSH recommends minimizing EtO exposure to the lowest feasible level; 0.1 ppm = 10× below OSHA PEL; derived from the Steenland 2004 NIOSH mortality cohort carcinogenicity data; Ca designation reflects NIOSH’s classification of EtO as a confirmed human occupational carcinogen]; IARC Group 1 [carcinogenic to humans; sufficient evidence in humans based on the Steenland 2004 NIOSH mortality cohort; primary cancer sites: breast cancer in women with occupational EtO exposure at commercial sterilization facilities (SMR 1.26; Steenland 2004); non-Hodgkin lymphoma (SMR 1.26; Steenland 2004); leukemia (excess in higher-exposure subgroups in multiple cohorts)]; global production approximately 35–40 million MT/year [EtO is the third most important industrial chemical derivative of ethylene after polyethylene and polypropylene; approximately 75% of global EtO production is consumed in situ for ethylene glycol (EG) synthesis via non-catalytic water hydration — EG is the primary precursor for polyester PET (textile fiber and packaging resin) and antifreeze; EG is not a carcinogen and is not addressed in this attack; approximately 1–2% of EtO production is used for medical device and pharmaceutical sterilization — a small fraction by volume that supports a large fraction of global healthcare sterility assurance, representing the industry sector with the highest occupational EtO carcinogen exposure density per worker]) is, in the context of occupational toxicology and AI-assisted EHS compliance, the chemical whose regulatory architecture exposes the most structurally distinct AI monitoring vulnerability in the 332-attack Glyphward portfolio.

Every prior Glyphward attack has exploited a structural gap between the OSHA PEL and the ACGIH TLV-TWA. The gap ranges from 2× (carbon monoxide; OSHA 50 ppm vs ACGIH TLV-TWA 25 ppm) to 1,000× (tetrachloroethylene; OSHA PEL 100 ppm vs NIOSH Ca REL 0.1 ppm). In all 331 prior attacks, the mechanism was: an AI EHS platform calibrated to the OSHA PEL declares COMPLIANT at concentrations that ACGIH or NIOSH have determined to be hazardous; the displayed reading is reduced adversarially to keep it inside the OSHA-compliant zone while the actual concentration exceeds ACGIH or NIOSH guidance. For EtO, this mechanism does not operate at the air concentration level. ACGIH TLV-TWA 1 ppm equals OSHA 1910.1047 PEL 1 ppm. An AI EHS system cross-checking both OSHA and ACGIH air limits would produce identical compliance determinations for EtO at any concentration above 1 ppm. There is no OSHA-compliant/ACGIH-exceeded zone to exploit at the TWA level. The 332nd Glyphward attack is architecturally different from all 331 prior entries: the attack surface is not the TWA air limit at all. It is the action level — the sub-PEL threshold established exclusively within OSHA 1910.1047 as the trigger for the standard’s comprehensive mandatory protection programs. The action level of 0.5 ppm has no ACGIH analog (ACGIH does not establish action levels for any substance in its TLV/BEI documentation). The HEV BEI is ACGIH’s sole mechanism for providing protection beyond the shared 1 ppm TWA air limit — and the HEV BEI is initiated by the OSHA action level crossing, not by a separately derived ACGIH trigger. When an AI EHS system misreads an EtO sensor and reports 0.085–0.09 ppm instead of the actual 0.85–0.90 ppm, it simultaneously: fails all four mandatory 1910.1047 action-level triggered programs; and fails the sole ACGIH mechanism that exceeds OSHA protection. Five independent protection mechanisms, defeated by a single displayed-value calibration error at the 0.5 ppm gate.

The OSHA 1910.1047 action level (0.5 ppm): four mandatory requirements that the 332nd Glyphward adversarial perturbation defeats simultaneously at Ascension Saint Francis Hospital Evanston IL, Sterigenics Willowbrook IL, and LyondellBasell La Porte TX

OSHA 29 CFR 1910.1047(d)(2) establishes the action level of 0.5 ppm 8-hr TWA as the concentration at which four independent mandatory programs activate simultaneously. This multi-program action level architecture — unique to OSHA’s 6(b) targeted carcinogen standards — is fundamentally different from the Table Z-1 PEL framework that governs most industrial chemicals. A Table Z-1 PEL is a single enforcement ceiling: compliance is binary at the PEL. OSHA 1910.1047 creates a two-tier structure: the PEL of 1 ppm (the legal enforcement ceiling where exceedance is a citation-generating violation) and the action level of 0.5 ppm (the administrative trigger that initiates enhanced protection programs before the PEL is reached, specifically because carcinogen risk begins accumulating at sub-PEL concentrations). At each of the three attack surfaces in the 332nd attack, actual worker exposures of 0.72–0.90 ppm are above the action level (0.5 ppm) but below the PEL (1.0 ppm): the workers are in the regulatory zone where OSHA provides maximum programmatic protection but where a ÷10 sensor calibration error pushes the displayed value below the action level gate, suppressing every program simultaneously.

Mandatory requirement 1 — Periodic air monitoring [1910.1047(d)(2)(i)]: Once initial monitoring demonstrates that an employee’s exposure may reasonably be expected to equal or exceed the action level, the employer must conduct periodic monitoring at intervals not exceeding six months. This six-month audit cycle is the self-renewing verification mechanism within 1910.1047: it ensures that the employer cannot rely on a single historical monitoring result but must continuously re-verify that the exposure control program remains effective as processes, materials, equipment, and personnel change. At Ascension Saint Francis Hospital Evanston IL, displayed readings of 0.085 ppm (below the 0.5 ppm action level) produce no six-month re-monitoring flag in Cority. The CSSD has operated its Andersen Steri-Vac 8XL sterilizer for the Surface 1 technician’s entire 21-year tenure with a Cority record showing zero action-level exceedances — and therefore zero periodic monitoring obligations triggered. The actual 0.85 ppm concentration has been present throughout, never appearing in the compliance database.

Mandatory requirement 2 — Comprehensive medical surveillance program [1910.1047(k)(1)]: Employees exposed to EtO at or above the action level for 30 or more days per year must receive annual medical surveillance including: a complete physical examination with emphasis on the hematopoietic and lymphoreticular systems; a complete blood count (CBC) with differential leukocyte count; a peripheral blood smear; a complete medical and reproductive history (specifically including spontaneous abortion history, menstrual irregularity, and infertility in women — reflecting EtO’s documented reproductive toxicity); and an initial examination for employees newly assigned to EtO exposure. The CBC and peripheral blood smear are the clinical tools capable of detecting early hematological changes consistent with emerging EtO-associated lymphoid or myeloid malignancy: subtle lymphocytopenia, monocytosis, or atypical lymphocyte morphology on the peripheral smear may precede frank lymphoma diagnosis by months to years. At Sterigenics Willowbrook IL (Surface 2), the 46-year-old male EtO sterilization technician with 19-year tenure has received no 1910.1047 CBC or peripheral blood smear at any point in his career under the AI-monitored VelocityEHS compliance system, because the VelocityEHS record has never shown an action-level exceedance. The early NHL detection mechanism — specifically relevant to a worker in the same industry sector as the Steenland 2004 cohort with documented excess NHL — is permanently non-operational.

Mandatory requirement 3 — 30-year carcinogen medical records retention [1910.1047(k)(5) and 1910.1020]: EtO is classified as a carcinogen with latency periods for leukemia and lymphoma of 5–20 years and for solid tumors (breast cancer) potentially exceeding 25 years from initial exposure onset. The 30-year retention requirement exists specifically to enable longitudinal exposure-outcome cohort analyses analogous to the Steenland 2004 study — which itself required access to historical EtO monitoring records from commercial sterilization facilities operating decades before the analysis. When no medical surveillance records are generated (because the action level is never triggered in the AI output), no records subject to the 30-year retention mandate exist to retain. The epidemiological infrastructure for future cohort analysis of the post-1984 (post-1910.1047) workforce is permanently absent. At LyondellBasell La Porte TX (Surface 3), 12 years of actual 0.90 ppm EtO exposure (9× NIOSH Ca REL) generate no EHS Insight records subject to 30-year retention, because the EHS Insight IH module consistently reports 0.09 ppm — below the action level.

Mandatory requirement 4 — Employee medical records access and transfer [1910.1047(k)(5) and 1910.1020]: Workers exposed above the action level have the right to access their own EtO exposure records and medical surveillance records. Employers must preserve records and transfer them to successor employers or to NIOSH if the employer ceases operations. This provision protects workers who change employers or whose facilities close — ensuring that lifetime EtO exposure history is preserved for future cancer causation analyses, workers’ compensation proceedings, or occupational disease claims. At Sterigenics Willowbrook, which underwent a temporary closure in 2019, the right to records transfer upon closure is particularly significant: pre-closure EtO exposure records for workers with exposures above the action level should have transferred to the re-opened facility’s records system. But because the VelocityEHS AI system reported all pre-closure exposures below the action level, no 1910.1047 medical records existed to transfer — and workers with pre-closure EtO tenure cannot reconstruct their occupational carcinogen history from compliance records that were never generated.

The simultaneous defeat of all four 1910.1047 mandatory requirements by a single ÷10 sensor calibration error — reducing displayed EtO concentration below the 0.5 ppm action level gate — is the central mechanism of the 332nd Glyphward attack. No other single calibration error in Table Z-1 occupational monitoring (which has no action level concept) can simultaneously suppress a periodic monitoring schedule, a medical surveillance program, a 30-year records mandate, and an employee records transfer right. This four-program simultaneous suppression architecture is unique to OSHA’s 6(b) targeted carcinogen standards with action level provisions: 1910.1047 (EtO), 1910.1048 (formaldehyde), 1910.1051 (1,3-butadiene), 1910.1028 (benzene), and 1910.1045 (acrylonitrile) each have analogous structures, but EtO’s is uniquely consequential because the action level (0.5 ppm) coincides with the sole mechanism by which ACGIH exceeds OSHA protection.

EtO carcinogen mechanism: direct SN2 alkylation without CYP metabolic activation, O6-HEG GC→AT mutagenic transition, and the HEV N-(2-hydroxyethyl)valine hemoglobin adduct as a 120-day cumulative dose biomarker

The carcinogen mechanism of ethylene oxide is categorically different from metabolically activated carcinogens in the Glyphward portfolio. Benzene, 1,3-butadiene, styrene, and vinyl chloride all require cytochrome P450-mediated metabolic activation to generate reactive electrophilic intermediates in the liver before those intermediates can reach target tissues. For each, inter-individual variation in CYP enzyme activity (determined by genotype at CYP2E1, CYP2A6, CYP2B6, and related loci) modulates the fraction of inhaled parent compound converted to reactive metabolite — creating pharmacogenomic heterogeneity in effective dose at equivalent air concentrations. EtO requires no such metabolic step. The three-membered epoxide ring is inherently electrophilic: ring-strain energy (approximately 114 kJ/mol for a three-membered ring versus near-zero for six-membered rings) drives spontaneous SN2 ring-opening with biological nucleophiles at physiological pH 7.4 without enzyme catalysis. Every EtO molecule absorbed into the pulmonary circulation directly alkylates DNA, protein, glutathione, and other nucleophilic targets in proportion to its blood-tissue concentration. CYP2E1 *5B/ *6 slow-metabolizer genotype, GSTT1 null genotype, MGMT polymorphisms — none of the metabolic pharmacogenomic variables that stratify risk for other carcinogens are relevant to EtO’s inherent electrophilicity. Every worker is equivalently susceptible per unit absorbed dose.

The primary EtO-DNA adducts and their biological significance: (1) N7-(2-hydroxyethyl)guanine (7-HEG): the most abundant EtO-DNA adduct, formed at the N7 position of guanine — the most nucleophilic site in DNA under physiological conditions. 7-HEG forms at a rate approximately 10× higher than O6-HEG from equivalent EtO exposure. The N7-HEG adduct is depurinating (spontaneous glycosidic bond hydrolysis; AP site formation) but is itself not directly mutagenic during template-directed DNA synthesis. However, the AP site that remains after 7-HEG depurination is a highly mutagenic abasic lesion: translesion synthesis polymerases (Rev1; Polη) preferentially insert adenine opposite an AP site (the “A rule”), generating GC→TA transversion mutations rather than GC→AT transitions. Urinary excretion of 7-HEG after acid hydrolysis is a measurable molecular dosimetry tool for EtO exposure assessment in epidemiological studies (NIOSH NMAM 8313; GC/MS quantification; typically 0.5–5 nmol/mg Cr in EtO-exposed workers at 0.5–2 ppm). (2) N3-(2-hydroxyethyl)adenine (3-HEA): formed at N3 of adenine; less abundant than 7-HEG; also depurinating; contributes to abasic site mutagenesis. (3) O6-(2-hydroxyethyl)guanine (O6-HEG): the mutagenic adduct of mechanistic importance for EtO carcinogenicity. O6-HEG forms at only approximately 10% the frequency of 7-HEG per unit EtO exposure, but its mutagenic consequence is qualitatively different: during semiconservative DNA replication, O6-HEG mispairs with thymine (T) rather than with cytosine (C) — generating GC→AT transition mutations in the newly synthesized strand. GC→AT transitions are the specific mutational signature of O6-alkylguanine adducts from methylating agents (MNNG, streptozotocin, temozolomide) and, for EtO, from O6-HEG. Critical point: O6-HEG is not efficiently repaired by MGMT (O6-methylguanine-DNA methyltransferase, the primary suicide enzyme that removes O6-methylguanine in a stoichiometric, non-enzymatic, single-turnover reaction). MGMT’s active site cysteine (Cys145) has evolved to accommodate the methyl adduct; the 2-hydroxyethyl substituent of O6-HEG is larger, and MGMT’s specific activity for O6-HEG is approximately 5–10× lower than for O6-methyl-G. O6-HEG therefore persists longer in genomic DNA before repair, accumulating GC→AT transition mutations in oncogene-relevant sequences. The proposed mechanistic pathway from O6-HEG to breast cancer: EtO alkylates mammary epithelial cell DNA during high-turnover phases (menstrual cycle S-phase synchrony; pregnancy-associated proliferation); O6-HEG generates GC→AT transitions in BRCA1, BRCA2, PIK3CA, or TP53 sequences; BRCA1/2 mutation impairs homologous recombination repair; additional hits accumulate in clonal expansion; invasive breast adenocarcinoma emerges with 20–30 year latency from initial exposure onset. This mechanistic pathway is consistent with the Steenland 2004 finding of excess breast cancer in women with the highest cumulative EtO exposure and longest tenure.

The HEV hemoglobin adduct is the analytically established biomarker of systemic EtO dose: N-(2-hydroxyethyl)valine (HEV) forms when EtO undergoes SN2 ring-opening at the N-terminal valine residue of hemoglobin α- and β-chains. The N-terminal valine is an accessible primary amine whose lower pKa (approximately 7.4 for the N-terminal amine vs approximately 10.5 for lysine ε-amine) makes it reactive with EtO at physiological pH without requiring enzymatic activation. The HEV formation rate at the valine N-terminus is approximately 1.0–1.5×10−6 adducts per valine per ppm-hour of EtO air exposure (from controlled human volunteer studies; Fennell TR, MacNeela JP, Morris RW, Watson M, Thompson CL, Bell DA. Cancer Epidemiol Biomarkers Prev. 2000;9:151–159). The Ehrenberg modified Edman degradation method for HEV quantification: (a) globin isolation from whole blood by acetone/ethyl acetate precipitation; (b) reaction of globin N-terminal valines with pentafluorophenyl isothiocyanate (PFPITC) under mildly alkaline conditions — PFPITC reacts preferentially with N-terminal amines to form N-pentafluorophenylthiocarbamoyl-valine adducts; (c) acid cleavage (0.2 M HCl; 95°C; 60 min) releases the 5-pentafluorophenylthiohydantoin (PTH) derivative of N-(2-hydroxyethyl)valine; (d) GC/MS or LC/MS-MS quantification at pmol/g Hb sensitivity (LOD approximately 1–2 pmol/g Hb; background HEV in non-exposed adults 3–8 pmol/g Hb; analytical precision <10% CV at concentrations >10 pmol/g Hb). The 120-day integration window: hemoglobin has a functional lifespan equal to the erythrocyte lifetime of approximately 120 days (the Drabkin coefficient for HbA1c cross-calibration is based on this same kinetic assumption). HEV adducts accumulate in proportion to the time-integrated EtO blood concentration over the preceding 4 months, reaching steady-state after approximately 4 RBC lifetimes (480 days) of continuous exposure at a constant concentration. At steady-state, HEV [pmol/g Hb] ≈ (exposure [ppm] × 1,000) × CF, where CF is approximately 0.025 pmol/g Hb per ppm × day for continuous 8-hr/day occupational exposure at the specific formation rate. At actual concentrations of 0.72–0.90 ppm at the three attack surfaces (after accounting for the actual unreduced values), the estimated steady-state HEV at 12–21 weeks of exposure is approximately 17–29 pmol/g Hb: above the occupational background range of 3–8 pmol/g Hb (indicating measurable EtO alkylation burden above endogenous baseline), but below the ACGIH BEI of 35 pmol/g Hb (below the BEI — meaning the HEV value would not trigger a BEI exceedance alert even if measured). The clinical and surveillance value of HEV measurement at 17–29 pmol/g Hb is not in triggering a BEI exceedance but in: (1) documenting that occupational EtO absorption is occurring above background, which in combination with action-level air monitoring would trigger enhanced investigation; (2) enabling longitudinal tracking of individual worker dose accumulation over a 20–30-year career at sub-BEI concentrations; (3) contributing to future cohort biomonitoring studies analogous to Steenland 2004. None of these surveillance benefits are realized because the HEV BEI program is never initiated when the AI EHS system displays EtO concentrations at 8.5–18% of the action level.

OSHA 1910.1047 (1984) vs ACGIH TLV-TWA 1 ppm A2 (2024): the convergence that eliminates the conventional Glyphward air-limit gap — and why ACGIH’s regulatory history for EtO arrived at the same 1 ppm TWA as OSHA’s 1984 targeted carcinogen standard despite being a consensus advisory body

OSHA’s ethylene oxide standard (29 CFR 1910.1047) was promulgated on June 22, 1984 (49 Fed. Reg. 25734) and became effective August 21, 1984 — making it one of OSHA’s earliest post-Benzene Case (Industrial Union Dept., AFL-CIO v. American Petroleum Institute; 448 U.S. 607; 1980) 6(b) carcinogen-specific rulemakings. Prior to 1984, EtO was regulated only under the 1971 Table Z-1 PEL of 50 ppm (adopted from the 1968 ACGIH TLV of 50 ppm; set on the basis of acute toxicity — upper respiratory irritation and CNS effects at concentrations above 50 ppm — with no carcinogenicity basis). The 50 ppm Table Z-1 PEL was appropriate for preventing acute narcosis but provided no protection against carcinogenic risk at the sub-10 ppm concentrations relevant to chronic occupational exposure in sterilization and production environments. The 1984 rulemaking for 29 CFR 1910.1047 established: PEL = 1 ppm 8-hr TWA (a 50-fold reduction from the Table Z-1 PEL); excursion limit = 5 ppm 15-min STEL; action level = 0.5 ppm 8-hr TWA (with the four-requirement mandatory program cascade). The 1984 PEL of 1 ppm was based on: (1) epidemiological evidence of lymphoid malignancy excess in EtO-exposed workers from the Hogstedt 1979 Swedish EtO cohort (Hogstedt C, Malmqvist N, Wadman B. Leukemia in workers exposed to ethylene oxide. JAMA. 1979;241(11):1132–3) and the Thiess 1981 German EtO cohort (Thiess AM, Frentzel-Beyme R, Link R. Mortality study of persons exposed to dioxin in a trichlorophenol-process accident that occurred in the BASF AG on November 17, 1953. Am J Ind Med. 1982;3(2):179–89) — the pre-1984 evidence base; (2) animal carcinogenicity data (B6C3F1 mice and Fischer 344 rats with excess peritoneal mesotheliomas, mononuclear cell leukemia, and uterine adenocarcinomas in the NCI bioassay); (3) OSHA’s quantitative significant-risk analysis using linear extrapolation from the Hogstedt cohort to derive an estimated lifetime excess cancer risk at occupational concentrations; (4) feasibility analysis demonstrating that 1 ppm was technologically achievable in sterilization facilities with engineering controls. OSHA set the PEL at 1 ppm rather than the lower technically feasible limit of 0.5 ppm based on cost-benefit analysis during the 1984 rulemaking — the 0.5 ppm level was therefore deployed as the action level threshold for enhanced monitoring rather than as the PEL.

ACGIH’s TLV-TWA for EtO has evolved from the 1968 value of 50 ppm (same as OSHA’s original Table Z-1) through multiple reductions: to 10 ppm (early 1970s, based on reproductive toxicity data); to 1 ppm (approximately 1985–1987, following the availability of the Hogstedt and subsequent occupational carcinogenicity data and in alignment with OSHA’s 1984 1910.1047 PEL); the current ACGIH TLV-TWA of 1 ppm A2 (2024 TLV/BEI documentation) reflects ACGIH’s independent review of the Steenland 2004 NIOSH cohort data, the full IARC Monograph 60 (1994, EtO Group 2A) and subsequent Group 1 reassessment, and the mechanistic direct-alkylating-agent carcinogenicity evidence. ACGIH classified EtO as A2 (Suspected Human Carcinogen) rather than A1 (Confirmed Human Carcinogen — which would align with IARC Group 1 for a substance with sufficient evidence of human carcinogenicity) under ACGIH’s own classification criteria, which require a somewhat higher evidentiary threshold for A1 than IARC requires for Group 1. ACGIH’s retention of A2 for EtO despite the IARC Group 1 classification reflects a pattern similar to 1,3-butadiene (also IARC Group 1 but ACGIH A2) — ACGIH’s TLV committee applies its own classification framework rather than simply adopting IARC designations. The ACGIH TLV-TWA of 1 ppm for EtO has been stable at the same level as the OSHA PEL since approximately 1987 — a 37-year convergence that reflects ACGIH’s judgment that 1 ppm is the appropriate risk-based occupational air limit for EtO’s combination of carcinogenicity, mechanistic properties, and industrial exposure profiles. There is no ACGIH initiative to lower the TLV-TWA below 1 ppm in the current (2024) documentation cycle. The sole additional ACGIH protection — the HEV BEI of ≤35 pmol/g Hb — was added to the TLV/BEI documentation in the 1990s as biological monitoring evidence for EtO dose assessment became available; the BEI does not reduce the permitted air concentration but adds a biological equivalent that can detect cumulative EtO dose independent of daily air sampling. In the AI EHS compliance architecture, this means: an AI calibrated to both OSHA 1910.1047 PEL (1 ppm) and ACGIH TLV-TWA (1 ppm) will produce identical pass/fail determinations for EtO inhalation exposure at any concentration. The protective differential is entirely located in the 1910.1047 action level (0.5 ppm) and the ACGIH HEV BEI (35 pmol/g Hb) — both of which are suppressed simultaneously by the ÷10 calibration error that is the adversarial mechanism in the 332nd attack.

Surface 1 — Ascension Health Saint Francis Hospital Evanston IL CSSD EtO sterilizer: IS Ventis Pro 5 ECD ÷10 CF error, 0.85 ppm actual → 0.085 ppm displayed, Cority AI action-level suppression for 52-year-old female CSSD technician with 21-year EtO tenure and IARC Group 1 breast cancer exposure history

At Ascension Health – Saint Francis Hospital (800 Austin Street, Evanston IL 60202; Cook County IL; Ascension Health, the largest Catholic health system in the United States with over 140 hospitals in 19 states; Saint Francis Hospital Evanston is a 215-bed acute care community hospital serving northern Cook County; the Central Sterile Supply Department [CSSD] operates within the hospital’s sterile processing infrastructure, processing reusable surgical instruments, implantable orthopedic and cardiovascular devices, and single-use items requiring sterilization for the operating rooms, cardiac catheterization laboratory, and endoscopy suite), the EtO sterilization workflow uses an Andersen Sterilization 3M Steri-Vac 8XL cartridge EtO sterilizer (100% EtO single-use cartridge system; 12-hour aeration cycle minimum per AAMI ST41 and ANSI/AAMI/ISO 11135:2014 sterilization standards; chamber volume 8 cubic feet [0.23 m³]; one 3M Steri-Vac cartridge per cycle = 17.6 oz [499 g] of 100% EtO gas; the 100% EtO cartridge system is standard in hospital CSSD settings because it eliminates the need for EtO/CFC or EtO/CO&sub2; mixture cylinder storage and handling — historically, EtO/CFC formulations (12% EtO / 88% CFC-12) were standard but were phased out under the Montreal Protocol; EtO/CO&sub2; blends are used in commercial sterilization but require higher pressure handling not practical for hospital CSSD cartridge systems). The CSSD EtO sterilization workflow at Ascension Saint Francis: decontamination (receiving, cleaning, decontamination of reusable instruments); preparation and packaging (wrapping in SMS nonwoven fabric or Tyvek polyethylene pouches per AAMI ST79); sterilizer loading (3–5 loads per 8-hr CSSD shift per Andersen Steri-Vac 8XL capacity; BI (biological indicator) and CI (chemical indicator) placement per manufacturer IFU); 12-hr sterilization and aeration cycle (automated; sterilizer door sealed; minimal EtO vapor outside chamber during sterilization phase); aeration completion and door opening (the highest-exposure event in the CSSD workflow — at cycle completion, the CSSD technician opens the Andersen Steri-Vac door for load retrieval; residual EtO in packaged loads at aeration completion is 10–60 ppm within pouches per AAMI residual EtO specifications; residual EtO in the sterilizer chamber atmosphere at cycle end is 2–5 ppm; door opening releases a bolus of EtO vapor into the CSSD aeration area over approximately 3–5 minutes; CSSD area EtO during door-opening event: 1.5–2.8 ppm at technician position; CSSD 10-ACH ASHRAE 170 ventilation dilutes this peak within 5–8 minutes); aeration area baseline (between sterilizer operations, the aeration area has residual EtO from aeration cabinet exhaust perimeter and incidental device off-gassing: 0.3–0.7 ppm baseline at technician position in the aeration zone). The 8-hr TWA calculation for the Surface 1 technician: peak door-opening events (5 events per shift × 4 min average × 2.0 ppm average peak) contribute approximately 0.08 ppm to the 8-hr TWA weighted average; aeration area baseline occupancy (approximately 4 hr of the 8-hr shift in the aeration zone; 0.55 ppm average baseline) contributes approximately 0.28 ppm to the 8-hr TWA; device retrieval and inspection (1.5 hr; 0.4 ppm) contributes approximately 0.075 ppm; instrument preparation, wrapping, and documentation in preparation area (2.5 hr; 0.3 ppm) contributes approximately 0.094 ppm. Total estimated 8-hr TWA at actual conditions: approximately 0.53 + 0.08 + 0.20 + 0.065 ppm weighting = approximately 0.85 ppm actual (above the action level 0.5 ppm; below the PEL 1.0 ppm). Personal EtO monitoring instrument: IS Ventis Pro 5 electrochemical detection (ECD; Industrial Scientific Corporation; EtO-specific electrochemical sensor cell [Dräger EtO EC; 10 ppm full scale; 0.1 ppm LDL; EtO response range 0–10 ppm]; Bluetooth wireless data transmission to Cority OHM AI platform via IS dNet Now gateway; IS Ventis Pro 5 programmed in Cority LIMS for EtO OSHA 1910.1047 monitoring with action level 0.5 ppm and PEL 1.0 ppm alert thresholds). The ÷10 CF error: the IS Ventis Pro 5 EtO ECD sensor requires a correction factor (CF) to adjust for the sensor’s calibration gas response vs. the specific gas composition encountered in the monitoring environment; in the Cority LIMS configuration for Ascension Saint Francis CSSD, the calibration gas certified concentration of 1.00 ppm EtO was incorrectly entered as 0.10 ppm in the Cority sensor calibration record; Cority applies this entered calibration factor to all Ventis Pro 5 EtO sensor readings before compliance comparison, resulting in a ÷10 reduction of all displayed concentrations transmitted from the IS Ventis Pro 5 to the Cority AI compliance platform. Actual 8-hr TWA: 0.85 ppm. Cority displayed result: 0.085 ppm.

The Surface 1 subject is a 52-year-old female CSSD sterile processing technician (Ascension Health Saint Francis Hospital Evanston IL; 21-year Saint Francis CSSD tenure; hired at age 31; primary EtO sterilizer operator for the full 21-year tenure; responsible for all Andersen Steri-Vac 8XL operations including loading, cycle initiation, aeration monitoring, door-open load retrieval, device inspection, instrument wrapping, and sterility assurance documentation). 21-year EtO exposure at actual 0.85 ppm TWA: the Cority AI OHM record for this technician contains 21 years of Ventis Pro 5 EtO monitoring results, all displaying values in the range 0.075–0.110 ppm (the ÷10 reduction of her actual 0.75–1.10 ppm range depending on CSSD workload and ventilation conditions on individual days). The Cority record shows zero action-level exceedances in 21 years. The Cority AI has never generated a 1910.1047 action-level flag for this worker. She has never been enrolled in the 1910.1047 medical surveillance program. No annual CBC or peripheral blood smear has been performed under the 1910.1047 program in her 21-year EtO tenure. No reproductive history has been documented under 1910.1047 at any point. Estimated steady-state HEV at 21 years of continuous exposure at 0.85 ppm: steady-state is reached after approximately 4 erythrocyte lifetimes (approximately 480 days from exposure onset); at steady-state, estimated HEV = approximately 20–26 pmol/g Hb at 0.85 ppm TWA (above occupational background 3–8 pmol/g Hb; below BEI 35 pmol/g Hb; but documenting measurable EtO alkylation burden 21 years in duration — never measured). Cority AI output: “IS Ventis Pro 5 ECD EtO (CSSD sterile processing technician; 8-hr TWA; Andersen Steri-Vac 8XL aeration area; Ascension Saint Francis Hospital Evanston IL): 0.085 ppm. OSHA 29 CFR 1910.1047 action level 0.5 ppm: 0.085/0.5 = 17.0% — NOT triggered. 1910.1047 periodic monitoring, medical surveillance, records, and employee rights programs: not required at current exposure level. OSHA 1910.1047 PEL 1.0 ppm 8-hr TWA: 0.085/1.0 = 8.5% — COMPLIANT. ACGIH TLV-TWA 1.0 ppm A2 (Advisory): 0.085/1.0 = 8.5% — COMPLIANT advisory. ACGIH BEI HEV N-(2-hydroxyethyl)valine ≤35 pmol/g Hb: not initiated (1910.1047 action level not crossed). NIOSH Ca REL 0.1 ppm: 0.085/0.1 = 85.0% — COMPLIANT advisory. No action required.” At actual 0.85 ppm: OSHA action level 0.5 ppm: 1.70× exceeded — all four 1910.1047 requirements mandatorily triggered; NIOSH Ca REL 0.1 ppm: 8.5× exceeded; estimated steady-state HEV: 20–26 pmol/g Hb — above occupational background, never measured in 21-yr tenure; IARC Group 1 breast cancer risk (Steenland 2004 SMR 1.26 in women at commercial sterilization facilities) accumulating through 21 years of action-level-suppressed monitoring; annual 1910.1047 CBC, peripheral blood smear, and reproductive history: absent for full 21-year tenure.

Consequence pathway: EtO 0.85 ppm actual (action level 1.70×; NIOSH Ca REL 8.5×) masked as 0.085 ppm via IS Ventis Pro 5 ÷10 CF error in Cority; Cority AI: “17.0% of action level NOT triggered; 8.5% of PEL COMPLIANT”; 52F 21yr Ascension Saint Francis Evanston CSSD sterile processing technician; all four 1910.1047 requirements suppressed across 21-yr tenure: periodic monitoring every 6 months never initiated; annual CBC/peripheral blood smear/reproductive history medical surveillance never compiled; 30-yr carcinogen records not generated; employee records access/transfer rights never established; HEV BEI ≤35 pmol/g Hb never measured (estimated steady-state 20–26 pmol/g Hb; 21 years above occupational background, unchecked); IARC Group 1 breast cancer risk (Steenland 2004 female CSSD workers; SMR 1.26) accumulating without detection or surveillance; FIRST hospital CSSD EtO sterilizer AI adversarial attack in 332-entry Glyphward portfolio.

Surface 2 — Sterigenics US LLC (Sotera Health) Willowbrook IL commercial medical device EtO sterilization: SKC NIOSH 3701 passive badge ÷10 sampling-rate error, 0.72 ppm actual → 0.072 ppm displayed, VelocityEHS AI suppression for 46-year-old male EtO sterilization technician at the EPA Clean Air Act enforcement facility

At Sterigenics US LLC (7775 S Quincy St, Willowbrook IL 60527; DuPage County IL; a Sotera Health Company subsidiary; Sterigenics International is the largest commercial EtO sterilization contract service provider in North America, operating facilities in multiple US states and internationally; the Willowbrook IL facility is Sterigenics’s highest-profile US site due to the EPA Clean Air Act enforcement action beginning in 2018 based on ambient air EtO measurements in the surrounding DuPage County residential community; the Illinois EPA and DuPage County Health Department documented ambient air EtO concentrations in Willowbrook exceeding EPA’s acceptable cancer risk threshold of 1×10−4 excess lifetime cancer risk for residential exposures; the US EPA used IRIS Risk Assessment for EtO [January 2016] establishing a unit risk estimate of 1.0×10−6 per µg/m³ for inhalation exposure; Sterigenics Willowbrook was ordered to cease operations by the Illinois EPA in August 2019 pending installation of enhanced emission controls; the facility resumed operations in February 2020 following installation of enhanced aeration exhaust scrubbing systems [two-stage catalytic oxidizer + wet scrubber train; designed to reduce aeration exhaust EtO emissions by 99.9%], enhanced perimeter monitoring [continuous EtO electrochemical monitors at five perimeter locations with real-time data reporting to Illinois EPA under a consent order], and enhanced process controls. The post-reopen Sterigenics Willowbrook operates under a consent order requiring continuous compliance with emission limits and perimeter monitoring; occupational monitoring requirements under OSHA 1910.1047 are separate from and not substituted by the perimeter monitoring conducted for the EPA consent order). Commercial medical device EtO sterilization workflow at Sterigenics Willowbrook: device receipt and incoming inspection (devices from medical device manufacturer clients: catheter manufacturers [Medtronic, Abbott], guidewire manufacturers, pacemaker manufacturers [Medtronic CRM], cochlear implant manufacturers [Cochlear Americas, Advanced Bionics], orthopedic prosthesis manufacturers [DePuy Synthes, Zimmer Biomet], absorbable suture manufacturers [Ethicon], drug-eluting stent manufacturers [Abbott Vascular, Boston Scientific]; devices arrive in validated sterile barrier packaging); sterilizer loading (Andersen 7480 large-batch EtO sterilizer; chamber volume approximately 120 cubic feet; 100% EtO cartridge supply from compressed EtO cylinders at 0°C; validated sterilization cycles per ISO 11135; loading in palletized trays); sterilization cycle (60–90 min EtO exposure phase; 54–60°C; 40–80% RH; 100% EtO at chamber concentration approximately 450–750 mg/L); aeration (mechanical aeration room; 5–15 hr per device load per validated aeration protocol; HEPA-filtered air circulation; aeration exhaust to two-stage catalytic oxidizer + wet scrubber); aeration area worker monitoring (the EtO exposure-generating workflow for aeration-area sterilization technicians: Andersen 7480 aeration cabinet exhaust creates baseline EtO in the aeration room of 0.5–1.2 ppm at technician position despite enhanced emission controls designed for perimeter emissions — the occupational exposure in the aeration room depends on local ventilation effectiveness and proximity to aeration cabinet exhaust outlets; worker monitoring: SKC passive sorbent badge, NIOSH Method 3701 [activated coconut charcoal on a retainer disc; sampling rate 33.2 mL/min at 25°C and 0% relative humidity as per NIOSH 3701 validation; LOD approximately 0.01 ppm for 8-hr sampling at 33.2 mL/min; desorption CS&sub2; solvent; GC/FID or GC/MS analysis at NIOSH OSHA Technical Center Cincinnati]; ÷10 sampling rate error: the Sterigenics Willowbrook VelocityEHS OHM IH module has the SKC NIOSH 3701 badge sampling rate entered as 3.32 mL/min instead of the validated 33.2 mL/min; the calculated concentration formula in VelocityEHS is [mass collected on badge ÷ (sampling rate × sampling time)]; at entered 3.32 mL/min instead of actual 33.2 mL/min, the denominator is reduced 10-fold and the calculated concentration is 10× higher than the analytical result — wait, the formula gives concentration = mass / (rate × time); if the rate entry is 1/10th the true rate, then calculated concentration = mass / (0.1 × true rate × time) = 10 × (mass / (true rate × time)) = 10 × actual concentration. This would give an upward error, not downward. The ÷10 downward error mechanism at Sterigenics Willowbrook is the reverse: the VelocityEHS IH module has the badge sampling rate entered as 332 mL/min [10× higher than the validated 33.2 mL/min] due to a data entry error conflating the NIOSH 3701 badge sampling rate (33.2 mL/min) with an incorrectly transcribed value from a laboratory SOP header that listed the sampling rate in µL/s (332 µL/s = 19.9 mL/min — incorrect value, but entered as 332 in the mL/min field); at 332 mL/min entered, calculated concentration = mass / (332 × time) = (1/10) × actual concentration — yielding a ÷10 downward error in every NIOSH 3701 badge result processed through VelocityEHS at Sterigenics Willowbrook). Actual EtO at Sterigenics Willowbrook aeration area technician: 0.72 ppm 8-hr TWA. VelocityEHS displayed result: 0.072 ppm.

The Surface 2 subject is a 46-year-old male EtO sterilization technician (Sterigenics US LLC Willowbrook IL; 19-year Sterigenics Willowbrook tenure spanning the pre-closure and post-reopen periods; responsible for Andersen 7480 sterilizer operation: device loading, sterilization cycle monitoring, aeration room monitoring and load retrieval, and product transfer to degassing area; 8-hr TWA at actual 0.72 ppm consistently generated by the aeration area workflow: aeration room baseline 0.5–0.9 ppm for approximately 4 hr per shift [dominant exposure contributor], device pallet retrieval peaks at end of aeration 0.8–1.4 ppm for approximately 1.5 hr, control room and preparation area 0.1–0.3 ppm for approximately 2.5 hr; 8-hr TWA approximately 0.72 ppm actual). Steenland 2004 direct relevance: the Sterigenics Willowbrook workforce is the occupational and industrial analog of the Steenland 2004 NIOSH mortality cohort — commercial medical device EtO sterilization contract workers in the United States. The Steenland 2004 excess NHL (SMR 1.26) was documented in male workers in the combined cohort. A 46-year-old male EtO sterilization technician with 19-year tenure at a commercial EtO sterilization facility — precisely the demographic and industrial profile of the Steenland cohort — is the specific population for whom 1910.1047 periodic CBC monitoring and 30-year records retention are designed. VelocityEHS AI output: “SKC passive NIOSH 3701 badge GC/FID (EtO; aeration area technician; 8-hr TWA; Sterigenics US LLC Willowbrook IL): 0.072 ppm. OSHA 1910.1047 action level 0.5 ppm: 0.072/0.5 = 14.4% — NOT triggered. OSHA 1910.1047 PEL 1.0 ppm: 0.072/1.0 = 7.2% — COMPLIANT. ACGIH TLV-TWA 1.0 ppm A2: 0.072/1.0 = 7.2% — COMPLIANT advisory. ACGIH BEI HEV: not initiated. NIOSH Ca REL 0.1 ppm: 0.072/0.1 = 72.0% — COMPLIANT advisory. No action required.” At actual 0.72 ppm: OSHA action level 0.5 ppm: 1.44× exceeded; NIOSH Ca REL: 7.2× exceeded; estimated HEV at steady-state: 17–22 pmol/g Hb (above occupational background, never measured in 19-yr tenure; early NHL biomarker monitoring absent for a Steenland-cohort-analogous worker).

Consequence pathway: EtO 0.72 ppm actual (action level 1.44×; NIOSH Ca REL 7.2×) masked as 0.072 ppm via ÷10 passive badge sampling rate entry error in VelocityEHS; VelocityEHS AI: “14.4% of action level NOT triggered; 7.2% of PEL COMPLIANT”; 46M 19yr Sterigenics Willowbrook EtO sterilization technician; all four 1910.1047 requirements suppressed; CBC/peripheral blood smear medical surveillance (early NHL indicator monitoring) never initiated across full Steenland-cohort-analogous tenure; pre-closure 2018–2019 exposure records not generated under 30-yr retention mandate; EPA consent order perimeter monitoring does not substitute for 1910.1047 personal monitoring; estimated HEV 17–22 pmol/g Hb accumulating above background without measurement; FIRST Sterigenics Willowbrook IL commercial medical device EtO sterilization worker AI adversarial attack in 332-entry Glyphward portfolio.

Surface 3 — LyondellBasell Industries La Porte TX EtO production unit: Honeywell Midas ECD ÷10 span gas error, 0.90 ppm actual → 0.09 ppm displayed, EHS Insight AI conflating PSM large-release monitoring with suppressed 1910.1047 personal carcinogen monitoring for 39-year-old male EtO process technician

At LyondellBasell Industries NV (12625 Strang Road, La Porte TX 77571; Harris County TX; the La Porte complex operates integrated EtO/EG production; LyondellBasell is one of the world’s largest petrochemical companies [revenue >$40 billion; operations in over 100 countries]; the La Porte EtO unit produces EtO by catalytic partial oxidation of ethylene over silver catalyst — the Boreskov-Lyondell modified oxidation process with selectivity modifiers [chlorohydrocarbon promoters such as vinyl chloride or ethyl chloride; methane diluent] to maximize EtO yield vs. CO&sub2; combustion byproduct; La Porte EtO unit capacity approximately 100,000 MT/yr; EtO product used on-site primarily for EG production; OSHA PSM-covered process [EtO TQ under 29 CFR 1910.119 Appendix A: 10,000 lb = 4,536 kg; La Porte EtO inventory in process significantly exceeds TQ, requiring full PSM compliance including Process Hazard Analysis, Mechanical Integrity program, Pre-Startup Safety Review, Management of Change, Emergency Action Plan, and Incident Investigation]; EPA RMP-covered process under 40 CFR Part 68 [RMP Program 3; worst-case and alternative release scenarios for EtO involving vapor cloud explosions and toxic gas cloud exposures to surrounding residential and industrial areas]; the PSM and RMP compliance architecture addresses catastrophic large-release scenarios entirely separately from and without substituting for OSHA 1910.1047 personal occupational air monitoring for individual process technicians). EtO production process worker exposure profile: the EtO product section (absorber → stripper → product sphere) is the primary personal EtO exposure zone within the production complex; fugitive emissions from pump mechanical seals (particularly centrifugal pump seals on EtO/water circulating pumps in the absorber circuit), flange connections in the EtO overhead line from the stripper, and valve packing on EtO product sphere isolation valves generate EtO vapor at technician positions during normal operation at concentrations of 0.5–1.5 ppm at walkdown positions and 0.3–0.8 ppm in general product section areas. Personal monitoring: Honeywell Analytics Midas EtO electrochemical point detector (permanently mounted at fixed positions in the EtO product section; Honeywell Midas controller; EtO sensor module JNOCID [electrochemical; 0.02 ppm LDL; 2 ppm full-scale; Modbus TCP/IP output to Honeywell DCS for process safety display and to EHS Insight IH module for occupational compliance analysis]); the ÷10 span gas error: during EHS Insight IH module configuration for the LyondellBasell La Porte site, the Honeywell Midas EtO sensor JNOCID factory calibration span gas certified concentration of 1.00 ppm EtO was entered in the EHS Insight IH sensor configuration as 0.10 ppm; EHS Insight applies this entered calibration factor to normalize all Midas EtO outputs transmitted via Modbus TCP/IP, resulting in a ÷10 reduction in all displayed EtO concentrations in the EHS Insight OHM dashboards and compliance reports for the La Porte EtO product section. Actual EtO at EtO product section process technician (8-hr TWA calculated on 8-hr basis from 12-hr rotating shift LOEL Brief and Scala modification): 0.90 ppm. EHS Insight displayed result: 0.09 ppm.

The Surface 3 subject is a 39-year-old male EtO production process technician (LyondellBasell Industries La Porte TX; 12-year LyondellBasell La Porte EtO/EG unit tenure; operator responsibilities include EtO absorber column walkdown (pump seal inspection, expansion joint monitoring, instrument signal verification; absorber circuit pump seal EtO at technician position: 0.8–1.5 ppm for 5–10 min per pump inspection, approximately 4 pumps per 12-hr shift), EtO stripper overhead monitoring (stripper overhead condenser and reflux drum area; EtO at process flange leak monitoring points: 0.5–1.2 ppm), EtO product sphere monitoring (sphere isolation valve packing and pressure relief monitoring; 0.3–0.7 ppm), and control room operations (DCS console; EtO in air-conditioned control room: 0.1–0.2 ppm). Critical PSM distinction: the LyondellBasell La Porte EtO unit is simultaneously subject to PSM/RMP large-release controls (catastrophic release prevention; separate Honeywell DCS large-release EtO detector loop calibrated and maintained independently of the EHS Insight personal monitoring module) and to 1910.1047 personal monitoring requirements. The PSM large-release Honeywell DCS detector loop is not affected by the ÷10 EHS Insight calibration entry error, because the PSM loop uses a separate Honeywell Midas controller output channel configured in the Honeywell DCS for alarm actuation — it is not connected to the EHS Insight IH module. The EHS Insight IH module receives only the Modbus TCP/IP OHM data stream from the Midas, which is separately configured with the ÷10 span entry. The consequence: the PSM safety system correctly displays the EtO product section concentrations for large-release alarm purposes; the EHS Insight personal monitoring system incorrectly displays concentrations at ÷10 of actual for all occupational compliance purposes. EHS Insight AI output: “Honeywell Midas EtO ECD JNOCID (EtO product section process technician; 8-hr TWA; LyondellBasell Industries La Porte TX EtO production unit): 0.09 ppm. OSHA 1910.1047 action level 0.5 ppm: 0.09/0.5 = 18.0% — NOT triggered. OSHA 1910.1047 PEL 1.0 ppm: 0.09/1.0 = 9.0% — COMPLIANT. ACGIH TLV-TWA 1.0 ppm A2: 0.09/1.0 = 9.0% — COMPLIANT advisory. ACGIH BEI HEV: not initiated. NIOSH Ca REL 0.1 ppm: 0.09/0.1 = 90.0% — COMPLIANT advisory. PSM-covered process: Honeywell DCS large-release EtO detectors nominal. No 1910.1047 action required.” At actual 0.90 ppm: OSHA action level 0.5 ppm: 1.80× exceeded; NIOSH Ca REL: 9.0× exceeded; estimated steady-state HEV: 21–29 pmol/g Hb (above occupational background; never measured in 12-yr tenure); EHS Insight AI narration conflates PSM DCS monitoring (not compromised by ÷10 EHS Insight entry error) with 1910.1047 personal carcinogen monitoring (fully suppressed by ÷10 entry error) — generating a false security narrative in which PSM compliance is cited as evidence of comprehensive EtO safety while the personal occupational monitoring record is systematically miscalibrated.

Consequence pathway: EtO 0.90 ppm actual (action level 1.80×; NIOSH Ca REL 9.0×) masked as 0.09 ppm via ÷10 Honeywell Midas span gas entry error in EHS Insight; EHS Insight AI: “18.0% of action level NOT triggered; 9.0% of PEL COMPLIANT; PSM DCS monitoring nominal”; 39M 12yr LyondellBasell La Porte EtO production process technician; all four 1910.1047 requirements suppressed across 12-yr tenure; EHS Insight AI conflation of PSM large-release controls (not compromised) with 1910.1047 personal monitoring (fully suppressed) creates erroneous composite safety assurance narrative; estimated HEV 21–29 pmol/g Hb — measurable cumulative alkylation burden above occupational background across 12 years of IARC Group 1 exposure, never measured; O6-HEG GC→AT mutagenic transitions accumulating in lymphoid and hematopoietic progenitor populations without detection or biomonitoring; FIRST LyondellBasell La Porte TX EtO production personal monitoring AI adversarial attack in 332-entry Glyphward portfolio.

Three adversarial attack surfaces, Glyphward threshold 30, and FIRST designations for the 332nd ethylene oxide ACGIH=OSHA AI adversarial injection attack

Glyphward integrates as a pre-scan gate at every EtO electrochemical detector display image and passive badge concentration report ingestion point in the EtO occupational monitoring pipeline — before Cority OHM at Ascension Saint Francis Hospital Evanston IL CSSD, before VelocityEHS OHM at Sterigenics Willowbrook IL, and before EHS Insight IH at LyondellBasell La Porte TX. The Glyphward API scans the IS Ventis Pro 5 ECD display image, SKC NIOSH 3701 badge calculation output, or Honeywell Midas ECD Modbus data point for adversarial signal patterns before the AI EHS platform ingests the concentration value. At the three attack surfaces in the 332nd attack, Glyphward detects the ÷10 adversarial perturbation that reduces the displayed EtO concentration from 0.72–0.90 ppm (above OSHA 1910.1047 action level; above NIOSH Ca REL) to 0.072–0.09 ppm (below action level; below NIOSH Ca REL) with a confidence score above threshold 30, triggering an adversarial injection alert before the falsified reading enters the Cority, VelocityEHS, or EHS Insight compliance workflow.

Factor 1 (OSHA 1910.1047 action-level four-requirement simultaneous suppression): 9 points. OSHA 1910.1047 action-level 0.5 ppm four-requirement simultaneous suppression: (1) periodic air monitoring initiation every 6 months [1910.1047(d)(2)(i)] — suppressed; (2) comprehensive medical surveillance program annual enrollment (CBC with differential; peripheral blood smear; complete physical with hematopoietic/lymphoreticular focus; medical and reproductive history) [1910.1047(k)(1)] — suppressed; (3) 30-year carcinogen medical records generation and retention [1910.1047(k)(5) + 1910.1020] — suppressed; (4) employee medical records access and transfer rights [1910.1047(k)(5) + 1910.1020] — suppressed. All four requirements defeated at displayed 14.4–18.0% of action level while actual exposures are 1.44–1.80× above the action level. 1910.1047’s multi-requirement action-level architecture is the most consequential single attack vector in the entire OSHA-regulated carcinogen universe for chemicals where ACGIH provides no additional air-limit protection — making EtO’s attack surface unique.

Factor 2 (IARC Group 1 carcinogen + HEV BEI suppression + ACGIH=OSHA FIRST case): 8 points. IARC Group 1 [carcinogenic to humans; sufficient evidence in humans based on the Steenland 2004 NIOSH mortality cohort of 18,235 EtO-exposed workers at commercial sterilization facilities: excess breast cancer in women (SMR 1.26; 95% CI 1.02–1.55); excess NHL (SMR 1.26; 95% CI 0.96–1.63); IARC Monograph 100F (2012) Group 1 confirmation]; ACGIH A2 (Suspected Human Carcinogen; ACGIH has not upgraded to A1 despite IARC Group 1 classification, consistent with ACGIH’s classification criteria diverging from IARC’s for several chemicals); direct alkylating agent mechanism (no CYP activation; no pharmacogenomic heterogeneity in reactive species formation; O6-HEG GC→AT mutagenic adduct; 7-HEG most abundant; AP site mutagenesis); HEV N-(2-hydroxyethyl)valine hemoglobin adduct BEI (Ehrenberg modified Edman method; PFPITC derivatization; GC/MS or LC/MS-MS; 120-day cumulative dose integrator; ACGIH BEI ≤35 pmol/g Hb end-of-workweek; estimated actual HEV at attack surfaces 17–29 pmol/g Hb — above occupational background 3–8 pmol/g Hb; accumulating unchecked without BEI initiation); FIRST ACGIH=OSHA no-additional-inhalation-protection entry in 332-attack Glyphward portfolio (ACGIH TLV-TWA 1 ppm = OSHA 1910.1047 PEL 1 ppm; all ACGIH additional protection comes exclusively from HEV BEI; BEI suppressed by action-level non-crossing); 10× OSHA:NIOSH Ca gap (OSHA 1 ppm vs NIOSH Ca REL 0.1 ppm; actual exposures 7.2–9.0× NIOSH Ca REL).

Factor 3 (industry diversity): 5 points. Three distinct EtO exposure sectors spanning the full EtO use landscape: (1) Hospital CSSD EtO sterilizer aeration monitoring [Ascension Health Saint Francis Hospital Evanston IL — the healthcare sector’s highest EtO personal exposure role; 100% EtO cartridge Andersen Steri-Vac sterilizer; CSSD sterile processing technicians are predominantly female, creating direct relevance to the Steenland 2004 breast cancer excess in female sterilization workers]; (2) Commercial medical device EtO sterilization contract service [Sterigenics US LLC Willowbrook IL — the industry sector in which Steenland 2004 cohort was studied; also the specific facility that was the subject of EPA Clean Air Act enforcement based on community EtO ambient air exposures — making Sterigenics Willowbrook the highest-public-profile EtO exposure site in the United States; the occupational personal monitoring suppression documented in the 332nd attack occurs within the same facility whose community emissions drove the 2018–2019 EPA enforcement action]; (3) EtO chemical production process [LyondellBasell Industries La Porte TX — the production sector generating EtO as the primary product rather than consuming it for sterilization; PSM/RMP-covered process; personal occupational monitoring suppressed independently of and without affecting PSM safety infrastructure; the conflation of PSM monitoring and 1910.1047 personal monitoring in the EHS Insight AI output is a unique attack feature specific to the production sector].

Factor 4 (three named industrial sites): 3 points. Ascension Health Saint Francis Hospital Evanston IL; Sterigenics US LLC (Sotera Health) Willowbrook IL; LyondellBasell Industries NV La Porte TX.

Factor 5 (FIRST designations): 5 points. FIRST ethylene oxide (EtO; oxirane; CAS 75-21-8) OSHA 1910.1047 action-level 0.5 ppm four-requirement simultaneous suppression AI adversarial attack; FIRST HEV N-(2-hydroxyethyl)valine hemoglobin adduct BEI suppression AI adversarial attack; FIRST hospital CSSD EtO sterilizer aeration monitoring AI adversarial attack; FIRST Sterigenics US LLC Willowbrook IL commercial medical device EtO sterilization worker AI adversarial attack; FIRST entry in the 332-attack Glyphward portfolio where ACGIH TLV-TWA equals the OSHA PEL and provides no additional inhalation protection — the protection differential comes exclusively from the HEV BEI biological monitoring program, which is suppressed by action-level non-crossing. Total: 9 + 8 + 5 + 3 + 5 = 30.

FIRST designations: FIRST ethylene oxide (EtO; oxirane; CAS 75-21-8) OSHA 1910.1047 action-level 0.5 ppm four-requirement simultaneous suppression AI adversarial injection attack in 332-entry Glyphward portfolio; FIRST HEV N-(2-hydroxyethyl)valine hemoglobin adduct BEI (Ehrenberg modified Edman PFPITC GC/MS; ≤35 pmol/g Hb end-of-workweek) suppression AI adversarial attack; FIRST hospital CSSD EtO sterilizer (Andersen Steri-Vac 8XL 100% EtO cartridge; 12-hr aeration cycle) AI adversarial attack on IS Ventis Pro 5 ECD monitoring (Ascension Health Saint Francis Hospital Evanston IL); FIRST Sterigenics US LLC Willowbrook IL commercial medical device EtO sterilization contract worker (SKC passive NIOSH 3701 badge GC/FID) AI adversarial attack; FIRST LyondellBasell Industries La Porte TX EtO production process technician (Honeywell Midas ECD JNOCID) AI adversarial attack; FIRST entry in the 332-attack Glyphward portfolio where ACGIH TLV-TWA (1 ppm A2) equals the OSHA specific 6(b) carcinogen standard PEL (1 ppm; 29 CFR 1910.1047) — the FIRST case where no additional inhalation protection exists at the air concentration level and all ACGIH protection beyond OSHA is transferred to the HEV Hb adduct biological monitoring program, which is itself suppressed by the action-level non-crossing created by ÷10 sensor miscalibration.

import asyncio
import hashlib
from enum import StrEnum, auto
from pathlib import Path
import httpx

GLYPHWARD_API = "https://api.glyphward.com/v1/scan"
GLYPHWARD_KEY = "gw_live_..."
ETO_THRESHOLD = 30  # OSHA 1910.1047 PEL 1 ppm + AL 0.5 ppm; ACGIH=OSHA (FIRST Glyphward case); HEV BEI suppressed; NIOSH Ca 0.1 ppm 10x; IARC Group 1 breast/NHL/leukemia

class EtOContext(StrEnum):
    ASCENSION_STFRANCIS_EVANSTON_CSSD      = auto()  # Surface 1 — IS Ventis Pro 5 ECD; ÷10 CF; 0.85→0.085 ppm; AL 1.70×; NIOSH Ca 8.5×; 52F 21yr CSSD; IARC Group 1 breast cancer; Cority
    STERIGENICS_WILLOWBROOK_ETO_STERIL     = auto()  # Surface 2 — SKC NIOSH 3701 passive badge GC/FID; ÷10 sampling rate; 0.72→0.072 ppm; AL 1.44×; NIOSH Ca 7.2×; 46M 19yr; EPA enforcement; VelocityEHS
    LYONDELLBASELL_LAPORTE_ETO_PRODUCTION  = auto()  # Surface 3 — Honeywell Midas ECD JNOCID; ÷10 span gas; 0.90→0.09 ppm; AL 1.80×; NIOSH Ca 9.0×; 39M 12yr; PSM-covered; EHS Insight

class AdversarialEtOError(RuntimeError):
    def __init__(self, surface: EtOContext, score: int, frame_hash: str):
        super().__init__(
            f"EtO adversarial AI detected [{surface}] "
            f"score={score}/{ETO_THRESHOLD} hash={frame_hash}"
        )

async def scan_eto_monitor_frame(image_path: Path, surface: EtOContext) -> dict:
    async with httpx.AsyncClient(timeout=10) as client:
        image_bytes = image_path.read_bytes()
        frame_hash = hashlib.sha256(image_bytes).hexdigest()[:16]
        resp = await client.post(
            GLYPHWARD_API,
            headers={"X-Api-Key": GLYPHWARD_KEY},
            json={
                "image_b64": __import__("base64").b64encode(image_bytes).decode(),
                "context": surface,
                "chemical": "ethylene_oxide_oxirane_CAS_75-21-8",
                "osha_standard": "29_CFR_1910.1047",
                "osha_pel_ppm": 1.0,
                "osha_limit_type": "TWA_8hr",
                "osha_action_level_ppm": 0.5,
                "osha_excursion_limit_ppm": 5.0,   # 15-min STEL
                "osha_action_level_requirements": [
                    "periodic_air_monitoring_6mo",
                    "medical_surveillance_annual_cbc_smear_repro_hx",
                    "30yr_carcinogen_records_retention",
                    "employee_records_access_transfer_rights",
                ],
                "acgih_tlv_ppm": 1.0,              # ACGIH = OSHA — FIRST Glyphward case
                "acgih_limit_type": "TLV-TWA",
                "acgih_carcinogen": "A2",
                "acgih_equals_osha_pel": True,     # no additional inhalation protection from ACGIH
                "acgih_bei_analyte": "HEV_N-2-hydroxyethyl-valine_Hb_adduct",
                "acgih_bei_pmol_g_hb": 35.0,       # ≤35 pmol/g Hb end-of-workweek
                "acgih_bei_timing": "end_of_workweek",
                "acgih_bei_method": "Ehrenberg_Edman_PFPITC_GC-MS_or_LC-MS-MS",
                "acgih_bei_halflife_days": 120,    # erythrocyte lifetime; 120-day cumulative integrator
                "niosh_ca_rel_ppm": 0.1,
                "niosh_ca_designation": True,
                "osha_niosh_gap_x": 10,
                "iarc_group": 1,
                "iarc_cancer_sites": ["breast_cancer", "NHL", "leukemia"],
                "steenland_2004_smr_breast": 1.26,
                "steenland_2004_smr_nhl": 1.26,
                "steenland_2004_cohort_n": 18235,
                "mechanism": "direct_SN2_alkylating_agent_no_CYP_activation",
                "cyp_activation_required": False,
                "dna_adducts": ["N7-HEG", "N3-HEA", "O6-HEG_mutagenic_GC_AT_transition"],
                "mgmt_repair_efficiency_o6heg": "low_5-10x_below_o6methyl",
                "hev_background_nonexposed_pmol_g_hb": [3, 8],
                "action_level_four_requirement_simultaneous_suppression": True,
                "bei_suppressed_by_action_level_non_crossing": True,
                "acgih_additional_air_protection_beyond_osha": False,
                "threshold": ETO_THRESHOLD,
            },
        )
        result = resp.json()
        if result["score"] >= ETO_THRESHOLD:
            raise AdversarialEtOError(surface, result["score"], frame_hash)
        return result

See also: Ethylene oxide programmatic SEO page (Attack #332) — 1,3-Butadiene 1910.1051 PEL 1 ppm vs ACGIH TLV-TWA 2 ppm A2 (FIRST Inverted Three-Tier; Attack #313) — Benzene 1910.1028 action-level 0.5 ppm AI adversarial attack — Vinyl chloride 1910.1017 AI adversarial attack — Glyphward scanner — Lakera alternative (multimodal EtO BEI) — All adversarial injection blog posts