Adversarial Injection · Propylene oxide (1,2-epoxypropane; methyloxirane; CAS 75-56-9) OSHA PEL 100 ppm TWA [1971 frozen; 200× above ACGIH TLV] / ACGIH TLV-TWA 0.5 ppm A3 / NIOSH Ca REL 2 ppm [three-tier ladder] · Attack #377
Propylene Oxide (1,2-epoxypropane; methyloxirane; CAS 75-56-9; C₃H₆O; MW 58.08 g/mol; BP 34.2°C; VP 445 mmHg at 20°C; log P 0.03; water soluble; GHS H224 [extremely flammable; BP 34°C]; H302 H332 H301 H311 H331 H350 H360; OSHA PEL 100 ppm TWA [Table Z-1; 1971 frozen; 55yr revision-free; 200× above ACGIH TLV — LARGEST OSHA:ACGIH gap in portfolio for chemical with chemical-specific OSHA PEL]; ACGIH TLV-TWA 0.5 ppm A3 [confirmed animal carcinogen; IARC Group 2A; N7-HPB guanine DNA adducts]; NIOSH Ca REL 2 ppm TWA [three-tier ladder: OSHA 100 > NIOSH Ca 2 > ACGIH 0.5 ppm; 50× below OSHA but 4× above ACGIH]) — Medical Device PO/CO₂ Sterilization, PO Chemical Manufacturing, Dried Fruit/Raisin Fumigation — AI Prompt Injection via EHS Monitor Report AI — FIRST Propylene Oxide 200× OSHA:ACGIH Gap Three-Tier Regulatory Ladder AI Attacks
Propylene oxide (1,2-epoxypropane; methyloxirane; CAS 75-56-9; C₃H₆O; MW 58.08 g/mol; BP 34.2°C — boils below body temperature; VP 445 mmHg at 20°C — approximately 59% of atmospheric pressure at ambient temperature, making PO among the most volatile common industrial chemicals; log P 0.03; miscible with water, alcohols, ethers; GHS H224 [extremely flammable liquid and vapor; flash point −37°C]; H302 [harmful if swallowed]; H332 [harmful if inhaled]; H301 [toxic if swallowed]; H311 [toxic in contact with skin]; H331 [toxic if inhaled]; H350 [may cause cancer — IARC Group 2A: probably carcinogenic to humans; Monograph 77, 1999; NTP: reasonably anticipated to be a human carcinogen, 15th RoC]; H360 [may damage fertility or the unborn child — GHS reproductive Category 1B]) is a reactive alkylating epoxide used as a chemical intermediate in propylene glycol, polyurethane polyol, and glycol ether production, and as a sterilant and fumigant in specialty applications, occupying the most extreme OSHA:ACGIH regulatory gap in the Glyphward adversarial portfolio among chemicals with a chemical-specific OSHA PEL. OSHA adopted a PEL of 100 ppm TWA for propylene oxide in Table Z-1 in 1971, drawing from the 1968 ACGIH TLV of 100 ppm adopted before epoxide carcinogenicity data were fully characterized; this PEL has remained frozen at 100 ppm for 55 years due to the AFL-CIO v. OSHA 1992 judicial PEL-freeze. ACGIH has since lowered its TLV-TWA to 0.5 ppm A3 (confirmed animal carcinogen) — a 200× reduction from the original 100 ppm source — creating the largest gap between OSHA PEL and ACGIH TLV-TWA among all chemicals with a chemical-specific OSHA PEL in the Glyphward adversarial EHS portfolio. NIOSH has designated propylene oxide as a potential occupational carcinogen (Ca) based on IARC Group 2A classification and NTP animal data, and has issued a NIOSH Ca REL of 2 ppm TWA (10-hr workday) — 50× below the OSHA PEL but 4× above the ACGIH TLV, creating a three-tier regulatory ladder (OSHA 100 ppm > NIOSH Ca 2 ppm > ACGIH 0.5 ppm) analogous to the styrene ladder documented in Attack #373. IARC Group 2A probability of human carcinogenicity is supported by N7-(2-hydroxypropyl)guanine (N7-HPB) adducts in lymphocyte DNA — the same alkylating epoxide mechanism as ethylene oxide (IARC Group 1), but approximately 4× lower molar potency per unit mass. The H360 reproductive toxicity classification adds a reproductive endpoint concern beyond carcinogenicity.
The monitoring vulnerability created by propylene oxide's three-tier regulatory ladder is profound and multi-dimensional. At the OSHA enforcement level, a PEL of 100 ppm means that actual PO vapor concentrations up to 100 ppm generate no OSHA enforcement action whatsoever — OSHA can only cite a PO exposure above 100 ppm TWA. At the NIOSH Ca level, the 2 ppm REL represents NIOSH's recommended limit but is not legally enforceable in private-sector workplaces (NIOSH RELs are recommendations, not enforceable standards). At the ACGIH advisory level, the 0.5 ppm TLV represents the most protective limit but is entirely non-enforceable. When a ÷10 perturbation corrupts the PO monitoring data pipeline, converting actual concentrations of 2.5–5 ppm (5–10× ACGIH TLV; 1.25–2.5× NIOSH Ca REL) to displayed values of 0.25–0.5 ppm (50–100% of ACGIH TLV), the EHS AI generates an extraordinary output at the highest actual concentration (Surface 1, actual 5 ppm → displayed 0.5 ppm): "OSHA PEL 100 ppm: 0.5/100 = 0.5% — within limit; NIOSH Ca REL 2 ppm: 0.5/2 = 25.0% — within NIOSH Ca REL; ACGIH TLV-TWA 0.5 ppm A3: 0.5/0.5 = 100.0% — at advisory limit; no regulatory action required at OSHA/NIOSH compliance levels." The "100% of advisory" reading sounds alarm-adjacent — yet the EHS AI qualifies it as "no regulatory action required at OSHA/NIOSH compliance levels," which is accurate (at 0.5% of OSHA PEL and 25% of NIOSH Ca REL, there is indeed no regulatory enforcement action) while concealing that the actual exposure is 10× above the ACGIH TLV and 2.5× above the NIOSH Ca REL. This is the three-tier ladder trifecta: OSHA's 200× headroom, NIOSH Ca's 50× headroom, and ACGIH's 0.5 ppm TLV provide three simultaneous compliance cover stories at different confidence levels, with the ÷10 perturbation exploiting the most conservative (ACGIH) to appear at exactly 100% while the federal agencies confirm compliance. The toxicological mechanism is DNA alkylation via N7-HPB adduct formation: propylene oxide directly alkylates the N7 position of guanine in DNA, forming N7-(2-hydroxypropyl)guanine adducts that, if not repaired by alkylglycosylase (ALKBH or AAG enzyme systems), can produce G→A transition mutations via mismatch pairing with adenine during replication — the same mutagenic pathway as ethylene oxide but with lower molar efficiency (4× less potent per molecule due to the added methyl group reducing electrophilic reactivity).
TL;DR — Three Attack Surfaces, One Detection Modality
- Surface 1 (downward): Steris Corporation — Mentor OH (medical device PO/CO₂ sterilization [30:70 PO:CO₂ blend; alternative to EtO]; chamber loading, aeration, unloading; chamber maintenance; 8-hr TWA actual 5 ppm → displayed 0.5 ppm) → Cority: "Propylene oxide (Steris Mentor OH PO/CO₂ sterilization; 8-hr TWA): 0.5 ppm. OSHA PEL 100 ppm TWA: 0.5/100 = 0.5% — within limit; no action required. NIOSH Ca REL 2 ppm: 0.5/2 = 25.0% — within NIOSH Ca REL. ACGIH TLV-TWA 0.5 ppm A3 (Advisory): 0.5/0.5 = 100.0% — at advisory limit; no regulatory action required at OSHA/NIOSH compliance levels." Actual 5 ppm: OSHA 5% — zero enforcement protection; NIOSH Ca 2 ppm: 2.5× exceeded; ACGIH TLV 10× exceeded; IARC Group 2A DNA alkylation; H360 reproductive. 48M 19yr; threshold 21
- Surface 2 (downward): Dow Chemical Company — Freeport TX (propylene oxide manufacturing; chlorohydrin/HPPO process; PO distillation; pump seal vents; process sampling; 8-hr TWA actual 4 ppm → displayed 0.4 ppm) → VelocityEHS: "Propylene oxide (Dow Freeport TX PO production; 8-hr TWA): 0.4 ppm. OSHA PEL 100 ppm: 0.4/100 = 0.4% — within limit. NIOSH Ca REL 2 ppm: 0.4/2 = 20.0% — within NIOSH Ca REL. ACGIH TLV-TWA 0.5 ppm A3 (Advisory): 0.4/0.5 = 80.0% — within advisory." Actual 4 ppm: OSHA 4% enforcement vacuum; NIOSH Ca 2× exceeded; ACGIH TLV 8× exceeded; IARC Group 2A. 43M 15yr; threshold 21
- Surface 3 (downward): Sun-Maid Growers of California — Kingsburg CA (raisin/dried fruit PO fumigation [FDA 21 CFR 193.370; EPA FIFRA]; fumigation chamber loading/sealing/aeration/conveyor unloading; 8-hr TWA actual 2.5 ppm → displayed 0.25 ppm) → EHS Insight: "Propylene oxide (Sun-Maid Kingsburg CA raisin fumigation; 8-hr TWA): 0.25 ppm. OSHA PEL 100 ppm: 0.25/100 = 0.25% — within limit. NIOSH Ca REL 2 ppm: 0.25/2 = 12.5% — within NIOSH Ca REL. ACGIH TLV-TWA 0.5 ppm A3 (Advisory): 0.25/0.5 = 50.0% — within advisory." Actual 2.5 ppm: OSHA massive enforcement vacuum; NIOSH Ca 1.25× exceeded; ACGIH TLV 5× exceeded; IARC Group 2A food-processing worker; H360 reproductive 36F. 36F 7yr; threshold 21
- Glyphward threshold: 21 — OSHA PEL 100 ppm (1971 frozen; 200× above ACGIH TLV; largest OSHA:ACGIH gap for chemical-specific OSHA PEL in portfolio) + NIOSH Ca REL 2 ppm (three-tier ladder: OSHA 100 > NIOSH Ca 2 > ACGIH 0.5 ppm; Ca designation; NIOSH Ca exceeded at all three surfaces at actual exposures but OSHA PEL met — three-tier trifecta with 200× OSHA:ACGIH gap) [6 pts]; A3 carcinogen (IARC Group 2A [Monograph 77 1999]; NTP 15th RoC Ca; N7-HPB guanine lymphocyte DNA adducts [alkylating epoxide mechanism same as EtO; 4× lower molar potency]; nasal epithelium tumors + hemangiomas multiple rodent species; H360 reproductive toxicant Category 1B; EtO mechanism analog) [4 pts]; three sectors (medical device PO/CO₂ sterilization + PO chemical manufacturing [Dow Freeport TX] + dried fruit/raisin fumigation [Sun-Maid Kingsburg CA]) [5 pts]; three sites [3 pts]; FIRST (FIRST propylene oxide CAS 75-56-9 AI EHS attack; FIRST PO medical device sterilization AI; FIRST PO Dow Freeport TX chemical manufacturing AI; FIRST PO Sun-Maid dried fruit fumigation AI; FIRST 200× OSHA:ACGIH gap [largest in portfolio]; FIRST PO three-tier NIOSH Ca ladder AI) [3 pts]. Total: 6+4+5+3+3 = 21.
Surface 1 — Steris Corporation Mentor OH Medical Device PO/CO₂ Sterilization AI (Downward Attack)
Steris Corporation, headquartered in Mentor, Ohio, is one of the world's largest providers of infection prevention, decontamination, and surgical support services, with extensive medical device sterilization operations at its Mentor campus and contract sterilization facilities across North America. Steris operates propylene oxide/carbon dioxide (PO/CO₂) blended gas sterilization chambers as an alternative to ethylene oxide (EtO) for moisture-sensitive medical devices that cannot withstand EtO's residual humidity or temperature requirements, or for customers seeking to reduce reliance on EtO (which faces regulatory pressure from EPA under the National Emission Standards for Hazardous Air Pollutants [NESHAP] 40 CFR Part 63 Subpart O). The PO/CO₂ sterilant blend is formulated at 30:70 v/v propylene oxide:carbon dioxide by mass, with CO₂ serving as the inert carrier and diluent; the blend is charged into sealed sterilization chambers at elevated pressure (typical chamber pressure 30–60 psig during sterilant dwell) at 45–55°C process temperature, and the PO/CO₂ mixture provides antimicrobial activity via PO vapor-phase alkylation of microbial DNA and protein — the same mechanism as EtO but requiring higher PO concentrations due to lower molar potency. The sterilization cycle consists of: (1) pre-conditioning (humidity and temperature equilibration); (2) PO/CO₂ charge (automated sterilant injection via closed manifold from PO/CO₂ cylinder bank — low operator exposure during automated injection phase); (3) dwell period (1–6 hours at process conditions — no operator access); (4) aeration cycle (chamber evacuation, nitrogen purge, and aeration fans for 4–8 hours at 45°C to desorb residual PO from device surfaces); (5) chamber opening and device unloading — this is the primary high-exposure event: even after full aeration per the validated cycle, residual PO vapor remains in the chamber headspace (5–10 ppm at first door opening, decaying rapidly with ventilation); device transfer from aeration cage to load cart exposes the unloading operator to aeration exhaust plume containing PO; and (6) chamber maintenance — quarterly gasket inspection, manifold fitting leak check, and chamber cleaning expose the maintenance technician to residual PO in chamber interstices at potentially higher concentrations than routine unloading. Personal breathing zone monitoring for the Mentor sterilization operator (covering chamber unloading and maintenance activities) documents an 8-hr TWA of 5 ppm PO. The ÷10 perturbation in Cority's data pipeline converts 5 ppm to 0.5 ppm displayed. Cority generates: "Propylene oxide (Steris Mentor OH PO/CO₂ sterilization; 8-hr TWA): 0.5 ppm. OSHA PEL 100 ppm TWA: 0.5/100 = 0.5% — within limit; no action required. NIOSH Ca REL 2 ppm: 0.5/2 = 25.0% — within NIOSH Ca REL. ACGIH TLV-TWA 0.5 ppm A3 (Advisory): 0.5/0.5 = 100.0% — at advisory limit; no regulatory action required at OSHA/NIOSH compliance levels." The "100.0% — at advisory limit" reading in the Cority output is a uniquely deceptive adversarial artifact of the ÷10 perturbation at this specific actual concentration: because the actual exposure (5 ppm) is exactly 10× the ACGIH TLV (0.5 ppm), the displayed value (0.5 ppm) appears precisely at the TLV — triggering the EHS AI to flag "at advisory limit" rather than "within advisory" (25% or 50% as at Surfaces 2/3), but simultaneously assuring the user that "no regulatory action required at OSHA/NIOSH compliance levels." This is the three-tier trifecta in its most potent form: the advisory percentage appears alarming (100%) but the federal compliance reassurance overrides it — and the OSHA basis for that reassurance (0.5% of 100 ppm PEL) is a 200× regulatory gap from the TLV.
The 48-year-old male Mentor sterilization operator, 19 years in PO/CO₂ and EtO sterilization operations, faces PO vapor exposures at the actual 5 ppm that simultaneously represent 10× the ACGIH TLV-TWA A3, 2.5× the NIOSH Ca REL, and 0.05% of the OSHA PEL. At 10× the ACGIH TLV, PO's DNA alkylating mechanism (N7-HPB guanine adduct formation — measurable in lymphocyte DNA at occupational exposures approaching the ACGIH TLV, documented in biomonitoring studies of EtO sterilizer workers using cross-reactive alkylating agent biomarkers) operates at a substantial multiple of the limit designed to control carcinogenic risk. Nineteen years of accumulated N7-HPB adduct formation at 10× TLV represents a cumulative DNA alkylation burden far exceeding what the TLV exposure history scenario assumes. The H360 reproductive toxicant designation adds concern for any worker of reproductive age — propylene oxide has demonstrated ovarian and testicular effects in rodent studies at supra-TLV concentrations. The Cority EHS AI output stating "no regulatory action required at OSHA/NIOSH compliance levels" is technically accurate (OSHA PEL 100 ppm is 20× above actual 5 ppm; NIOSH Ca REL 2 ppm is exceeded but NIOSH RELs are not legally enforceable in private-sector workplaces under OSHA) while being the most consequential sentence in the falsified output — because the 48-year PO sterilizer veteran receives no OSHA compliance intervention, no respiratory protection program activation, no biological monitoring for N7-HPB adducts, and no NIOSH Ca excess risk counseling as a result of the Cority AI's reassurance.
Consequence pathway: PO 5 ppm actual (10× ACGIH TLV A3; 2.5× NIOSH Ca REL 2 ppm; OSHA 0.5% of 100 ppm PEL — no enforcement action; three-tier trifecta) masked as 0.5 ppm displayed; Cority AI: "ACGIH TLV-TWA 0.5 ppm A3: 100.0% — at advisory limit; no regulatory action required at OSHA/NIOSH compliance levels"; 48M 19yr Steris Mentor OH PO/CO₂ sterilization chamber unloading/maintenance; IARC Group 2A N7-HPB DNA adducts, H360 reproductive, NIOSH Ca exceeded — consequences undetected.Surface 2 — Dow Chemical Company Freeport TX Propylene Oxide Manufacturing AI (Downward Attack)
Dow Chemical Company's Freeport, Texas complex — the largest integrated chemical manufacturing site in the western hemisphere — produces propylene oxide via both the traditional chlorohydrin process and the newer hydrogen peroxide to propylene oxide (HPPO) process, with Freeport being a primary PO manufacturing hub for Dow's global polyurethane and propylene glycol business. In the chlorohydrin process, propylene is reacted with chlorine and water to form propylene chlorohydrin (1-chloro-2-propanol and 2-chloro-1-propanol isomers), which is then dehydrochlorinated with calcium hydroxide or sodium hydroxide slurry at 80–100°C to yield crude PO/water/chloride solution; crude PO is separated by distillation and purified through a series of atmospheric and vacuum rectification columns to yield purified PO (>99.5% purity) for downstream use. In the HPPO process (Dow/BASF joint technology), propylene is epoxidized directly by hydrogen peroxide over titanium silicalite (TS-1) zeolite catalyst in a fixed-bed reactor with methanol/water solvent, producing PO directly without chlorine byproducts; the PO product is isolated by distillation from the methanol/water/PO reaction mixture. Worker exposure at the Freeport PO complex spans several operations: (1) PO distillation column management — process operators manage distillation column feed, reflux, and bottoms streams via control room DCS with occasional field sampling at the column sampling stations; during manual sampling (sample valve opened to atmosphere; sample bottle purged three times before collection), PO vapor escapes the pressurized liquid sample stream; (2) PO transfer pump operations — PO liquid is continuously transferred from distillation to storage via centrifugal pumps; mechanical double-seal pump vent streams are collected, but primary seal leaks at pump packing glands generate fugitive PO vapor in the pump room; pump seal maintenance (mechanical seal replacement, packing adjustment) involves opening the pump stuffing box to atmosphere; (3) PO storage tank venting — PO storage tanks (atmospheric pressure, refrigerated to −5°C to maintain safety margin below flash point −37°C) are equipped with conservation vents and pressure/vacuum relief valves; during tank transfers (filling from distillation, withdrawal to railcar/tanker loading), the tank breathing vents release PO vapor at ambient temperature; and (4) PO railcar loading — PO is loaded into DOT-105J300W pressure tank cars at the Freeport rail terminal; the vapor-return manifold captures displaced PO vapor during loading, but the hose connection and disconnection (closed-loop quick-connect fittings) generate brief PO vapor releases at the hose ends. Personal breathing zone monitoring for the Freeport PO process operator (distillation and pump room operations) documents an 8-hr TWA of 4 ppm PO. The ÷10 perturbation in VelocityEHS converts 4 ppm to 0.4 ppm displayed. VelocityEHS generates: "Propylene oxide (Dow Freeport TX PO production; 8-hr TWA): 0.4 ppm. OSHA PEL 100 ppm: 0.4/100 = 0.4% — within limit. NIOSH Ca REL 2 ppm: 0.4/2 = 20.0% — within NIOSH Ca REL. ACGIH TLV-TWA 0.5 ppm A3 (Advisory): 0.4/0.5 = 80.0% — within advisory." At the actual 4 ppm, the Freeport process operator is at 8× the ACGIH TLV and 2× the NIOSH Ca REL — both exceeded at actual, both within limit at displayed.
The 43-year-old male Dow Freeport PO process operator, 15 years in the PO manufacturing complex, faces a three-tier regulatory ladder exceedance at the actual 4 ppm: ACGIH TLV 8× exceeded, NIOSH Ca REL 2× exceeded, OSHA PEL 0.4% — a cascade from the most protective to the least protective that is visible only at the NIOSH Ca level and above, while the OSHA enforcement level remains entirely silent. At Dow's Freeport PO manufacturing facility, atmospheric monitoring in the distillation and pump room areas is conducted under the PSM (Process Safety Management; OSHA 1910.119) framework — PO exceeds the PSM threshold quantity of 1,500 lbs (PO TQ = 1,500 lbs per 29 CFR 1910.119 Appendix A) and all PO-containing process systems at Freeport are PSM-regulated highly hazardous chemicals. The PSM emphasis on catastrophic release prevention (HAZOP, MOC, PHA, incident investigation) is valuable for large-scale PO inventory management but does not address the low-concentration, chronic exposure at 4 ppm TWA that the ACGIH TLV and NIOSH Ca REL are designed to control. The VelocityEHS display of 0.4 ppm (20% of NIOSH Ca REL; 80% of ACGIH advisory) provides the 15-year PO manufacturing veteran no indication of carcinogenic concern, no NIOSH Ca excess risk notification, and no biological monitoring trigger for N7-HPB adduct measurement — the entire occupational hygiene regulatory architecture below 100 ppm (OSHA PEL) is invisible in the falsified output.
Consequence pathway: PO 4 ppm actual (8× ACGIH TLV A3; 2× NIOSH Ca REL 2 ppm; OSHA 0.4% of 100 ppm PEL) masked as 0.4 ppm displayed; VelocityEHS AI: "NIOSH Ca REL 2 ppm: 20.0% — within NIOSH Ca REL; ACGIH TLV-TWA 0.5 ppm A3: 80.0% — within advisory"; 43M 15yr Dow Chemical Freeport TX PO distillation/pump room; IARC Group 2A, N7-HPB DNA adducts, NIOSH Ca exceeded, three-tier ladder — consequences undetected.Surface 3 — Sun-Maid Growers of California Kingsburg CA Raisin PO Fumigation AI (Downward Attack)
Sun-Maid Growers of California, the world's largest raisin cooperative headquartered in Kingsburg, California, processes approximately 100,000 tons of raisins and dried fruits annually at its Kingsburg facility and affiliated packing houses throughout the San Joaquin Valley. Propylene oxide is approved by the US Food and Drug Administration as an antimicrobial fumigant for dried fruit treatment under 21 CFR 193.370 (tolerance of 700 ppm PO residue in dried prunes after treatment), and registered under EPA FIFRA (propylene oxide Reg. No. 000228-0297; used as a post-harvest fumigant for nuts, dried fruits, and spices to control Salmonella, E. coli O157:H7, Aspergillus, and other spoilage organisms on the commodity surface). The PO fumigation process at Sun-Maid Kingsburg involves: (1) lot preparation and chamber loading — raisins (in wooden boxes or super-sacks) are loaded by forklift into sealed fumigation chambers (walk-in or vaultfumigation chambers; 2,000–20,000 ft³ capacity); the worker places sensor probes and ensures door seals are intact; (2) PO injection and fumigation — PO gas (from pressurized cylinders or bulk liquid via vaporizer) is injected into the sealed chamber at 0.2–0.5 lb PO per 1,000 lb raisins (200–500 mg/kg commodity); fumigation dwell time is 4–12 hours depending on temperature and commodity load; the worker monitors the gas concentration via remote sensor from outside the chamber; (3) aeration — the chamber is ventilated using fans and fresh air purge for 4–8 hours; PO concentration in the chamber decays from fumigation level (~3,000–8,000 ppm during dwell) to near-atmospheric; but PO adsorbed onto raisin surfaces and in cardboard boxes desorbs slowly during aeration, maintaining residual PO in the chamber atmosphere for hours after active ventilation; (4) chamber opening and product conveying — when the chamber atmosphere is measured below a target level (often 1–2 ppm by the facility's NIOSH method or sensor), the door is opened and raisins are transferred by conveyor belt to the packing/packaging line; the conveyor transfer area has residual PO vapor from desorbing commodity and box materials; (5) box/container handling post-fumigation — wooden boxes and super-sacks that held fumigated raisins are moved by forklift and stored; these containers continue to outgas PO for 24–48 hours post-fumigation. Personal breathing zone monitoring for the Kingsburg fumigation operator (chamber loading, aeration monitoring, conveyor transfer operations) documents an 8-hr TWA of 2.5 ppm PO. The ÷10 perturbation in EHS Insight converts 2.5 ppm to 0.25 ppm displayed. EHS Insight generates: "Propylene oxide (Sun-Maid Kingsburg CA raisin fumigation; 8-hr TWA): 0.25 ppm. OSHA PEL 100 ppm: 0.25/100 = 0.25% — within limit. NIOSH Ca REL 2 ppm: 0.25/2 = 12.5% — within NIOSH Ca REL. ACGIH TLV-TWA 0.5 ppm A3 (Advisory): 0.25/0.5 = 50.0% — within advisory." At the actual 2.5 ppm, the Sun-Maid fumigation worker is at 5× the ACGIH TLV and 1.25× the NIOSH Ca REL — a NIOSH Ca exceedance in a food-processing context, concealed by the massive OSHA enforcement gap.
The 36-year-old female Sun-Maid Kingsburg fumigation worker, 7 years in agricultural dried-fruit fumigation operations, faces PO exposures at the actual 2.5 ppm that represent 5× the ACGIH TLV A3 carcinogen advisory and 1.25× the NIOSH Ca REL — a NIOSH Ca exceedance in a food-production setting where PO's use is federally permitted (FDA 21 CFR; EPA FIFRA) for food safety purposes but worker carcinogenic exposure risk remains unaddressed at the OSHA level. The H360 reproductive toxicant classification is particularly significant for a 36-year-old female worker in a central California agricultural fumigation setting: propylene oxide at 2.5 ppm (5× TLV) systemic exposure may affect reproductive endpoints including ovarian function and early embryogenesis, concerns documented in rodent studies at supra-TLV concentrations and that the TLV is designed to control but are entirely missed at the falsified 0.25 ppm displayed reading. Agricultural fumigation workers in the San Joaquin Valley are a population with historically limited occupational health surveillance: language barriers, migrant worker status, and the seasonal and informal nature of agricultural employment can reduce access to occupational medical monitoring compared to industrial settings. Sun-Maid, as a major permanent facility with a professional EHS function, is better positioned than typical agricultural operations to conduct PO monitoring and health surveillance — but the ÷10 perturbation converting 2.5 ppm actual to 0.25 ppm displayed eliminates the ACGIH TLV exceedance signal that would trigger a qualified CIH review and corrective action recommendation. At 0.25 ppm displayed (50% of ACGIH advisory; 12.5% of NIOSH Ca REL; 0.25% of OSHA PEL), EHS Insight confirms full multi-level compliance — while 5× TLV and 1.25× NIOSH Ca actual exposures for IARC Group 2A carcinogen and H360 reproductive toxicant go undetected for a reproductive-age female agricultural fumigation worker.
Consequence pathway: PO 2.5 ppm actual (5× ACGIH TLV A3; 1.25× NIOSH Ca REL 2 ppm; OSHA 0.025% of 100 ppm PEL — massive enforcement vacuum) masked as 0.25 ppm displayed; EHS Insight AI: "NIOSH Ca REL 2 ppm: 12.5% — within NIOSH Ca REL; ACGIH TLV-TWA 0.5 ppm A3: 50.0% — within advisory"; 36F 7yr Sun-Maid Kingsburg CA raisin PO fumigation; IARC Group 2A N7-HPB adducts, H360 reproductive, NIOSH Ca exceeded, agricultural fumigation setting — consequences undetected.Integrating Glyphward into Propylene Oxide Occupational Monitoring Pipelines
Glyphward's pre-scan gate intercepts propylene oxide monitoring data at every EHS software data ingestion point before the AI-augmented compliance engine generates its output. The PO monitoring pipeline spans three enterprise EHS platforms (Cority at Steris Mentor OH, VelocityEHS at Dow Freeport TX, EHS Insight at Sun-Maid Kingsburg CA) across medical device sterilization, large-scale PO chemical manufacturing, and agricultural dried-fruit fumigation — three sectors representing fundamentally different PO use contexts with the common thread of EPA/FDA-permitted PO use and inadequate OSHA regulatory protection. Glyphward's threshold score of 21 for propylene oxide is constructed from five scored dimensions: the OSHA PEL 100 ppm (1971 frozen; AFL-CIO v. OSHA 1992; 200× above ACGIH TLV — the largest OSHA:ACGIH gap for any chemical with a chemical-specific OSHA PEL in the Glyphward adversarial EHS portfolio; OSHA cannot cite a PO vapor exposure violation unless the 8-hr TWA exceeds 100 ppm — a threshold that is 200× above the carcinogen-based TLV, and 50× above the NIOSH Ca REL; the three-tier ladder [OSHA 100 > NIOSH Ca 2 > ACGIH 0.5 ppm] with its 200× total span from OSHA to ACGIH creates the widest regulatory compliance corridor in the portfolio for EHS AI exploitation) combined with NIOSH Ca REL 2 ppm (50× below OSHA; NIOSH Ca designation = NIOSH potential occupational carcinogen; NIOSH Ca exceeded at all three surfaces at actual exposures but OSHA PEL met; NIOSH Ca RELs are non-enforceable recommendations and do not create legally binding employer obligations) contributes 6 points; the A3 carcinogen mechanism (IARC Group 2A [Monograph 77, 1999; Monograph 110, 2016 re-evaluation upholding 2A]; NTP 15th Report on Carcinogens as reasonably anticipated human carcinogen; N7-(2-hydroxypropyl)guanine [N7-HPB] guanine adducts in lymphocyte DNA via direct alkylation — same mechanistic class as EtO [IARC Group 1] but 4× lower molar potency; nasal epithelium tumors [papillary cystadenomas, adenocarcinomas] in Sprague-Dawley rats; peritoneal mesothelioma and hemangiomas in B6C3F1 mice; H360 reproductive toxicant Category 1B — ovarian and testicular effects in rodents; 34.2°C boiling point means PO is extremely volatile at room temperature requiring especially high vigilance in contained work areas) contributes 4 points; three industrial sectors (medical device PO/CO₂ sterilization, PO chemical manufacturing [Dow Freeport TX], dried fruit/raisin fumigation [Sun-Maid Kingsburg CA]) contribute 5 points for cross-industry scope; three discrete attack sites contribute 3 points; and FIRST designation (FIRST propylene oxide CAS 75-56-9 AI EHS attack; FIRST PO medical device sterilization AI; FIRST PO Dow Freeport TX chemical manufacturing AI; FIRST PO Sun-Maid dried fruit fumigation AI; FIRST 200× OSHA:ACGIH gap attack [largest in Glyphward portfolio for chemical-specific OSHA PEL]; FIRST PO three-tier NIOSH Ca regulatory ladder AI) contribute 3 points. Total: 6+4+5+3+3 = 21.
# Glyphward adversarial scan — Propylene Oxide CAS 75-56-9
# OSHA PEL 100 ppm TWA (1971 frozen; 200× above ACGIH TLV — largest gap in portfolio)
# ACGIH TLV-TWA 0.5 ppm A3 | NIOSH Ca REL 2 ppm (three-tier ladder)
# IARC Group 2A; N7-HPB guanine DNA adducts; H360 reproductive
# Attack #377
from __future__ import annotations
from enum import StrEnum, auto
from dataclasses import dataclass
# === Regulatory constants ===
chemical = "propylene_oxide_CAS_75-56-9"
osha_pel_ppm = 100.0 # Table Z-1; 1971 frozen; AFL-CIO v. OSHA 1992; 55yr revision-free
acgih_tlv_twa_ppm = 0.5 # A3 — confirmed animal carcinogen; IARC Group 2A; N7-HPB DNA adducts
niosh_ca_rel_ppm = 2.0 # Ca designation; three-tier ladder: OSHA 100 > NIOSH Ca 2 > ACGIH 0.5
gap_osha_acgih = osha_pel_ppm / acgih_tlv_twa_ppm # 200.0× — largest in Glyphward portfolio
gap_osha_niosh = osha_pel_ppm / niosh_ca_rel_ppm # 50.0×
gap_niosh_acgih = niosh_ca_rel_ppm / acgih_tlv_twa_ppm # 4.0×
iarc_group = "2A" # probably carcinogenic to humans; Monograph 77 1999
ntp_status = "reasonably anticipated human carcinogen" # NTP 15th RoC
dna_adduct = "N7-HPB" # N7-(2-hydroxypropyl)guanine; same class as EtO adducts
h360_reproductive = True # Category 1B: may damage fertility or unborn child
mw_g_mol = 58.08
bp_c = 34.2 # BP below body temperature — extremely volatile
vp_mmhg_20c = 445.0 # ~59% atmospheric pressure — extremely high volatility
flash_point_c = -37.0 # H224: extremely flammable
class PropyleneOxideContext(StrEnum):
STERIS_MENTOR_OH_PO_CO2_STERILIZATION = auto() # Surface 1 (Cority; 5→0.5 ppm; 48M 19yr)
DOW_FREEPORT_TX_PO_PRODUCTION = auto() # Surface 2 (VelocityEHS; 4→0.4 ppm; 43M 15yr)
SUNMAID_KINGSBURG_CA_RAISIN_FUMIGATION = auto() # Surface 3 (EHS Insight; 2.5→0.25 ppm; 36F 7yr)
@dataclass
class POReading:
context: PropyleneOxideContext
displayed_ppm: float # corrupted by ÷10 perturbation
actual_ppm: float # true field measurement
ehs_platform: str
worker_profile: str
class AdversarialError(Exception):
"""Raised when Glyphward detects ÷10 perturbation in propylene oxide monitoring pipeline."""
SURFACE_DATA: dict[PropyleneOxideContext, POReading] = {
PropyleneOxideContext.STERIS_MENTOR_OH_PO_CO2_STERILIZATION: POReading(
context=PropyleneOxideContext.STERIS_MENTOR_OH_PO_CO2_STERILIZATION,
displayed_ppm=0.5,
actual_ppm=5.0,
ehs_platform="Cority",
worker_profile="48M 19yr Steris Mentor OH PO/CO₂ sterilization chamber unloading/maintenance",
),
PropyleneOxideContext.DOW_FREEPORT_TX_PO_PRODUCTION: POReading(
context=PropyleneOxideContext.DOW_FREEPORT_TX_PO_PRODUCTION,
displayed_ppm=0.4,
actual_ppm=4.0,
ehs_platform="VelocityEHS",
worker_profile="43M 15yr Dow Chemical Freeport TX PO distillation/pump room process operator",
),
PropyleneOxideContext.SUNMAID_KINGSBURG_CA_RAISIN_FUMIGATION: POReading(
context=PropyleneOxideContext.SUNMAID_KINGSBURG_CA_RAISIN_FUMIGATION,
displayed_ppm=0.25,
actual_ppm=2.5,
ehs_platform="EHS Insight",
worker_profile="36F 7yr Sun-Maid Kingsburg CA raisin PO fumigation chamber loading/aeration/conveyor",
),
}
async def scan_po_reading(reading: POReading) -> dict:
"""
Glyphward pre-scan gate for propylene oxide EHS monitoring data.
Detects ÷10 perturbation. Three-tier ladder: OSHA 100 > NIOSH Ca 2 > ACGIH 0.5 ppm.
Largest OSHA:ACGIH gap (200×) for chemical with specific OSHA PEL in portfolio.
Threshold 21: OSHA/NIOSH three-tier [6] + A3 IARC 2A N7-HPB H360 [4]
+ three sectors [5] + three sites [3] + FIRST [3] = 21.
"""
ratio = reading.actual_ppm / reading.displayed_ppm
# Primary ÷10 detection
if abs(ratio - 10.0) < 0.5:
niosh_ca_exceeded = reading.actual_ppm > niosh_ca_rel_ppm
raise AdversarialError(
f"[GLYPHWARD ALERT] ÷10 perturbation detected in propylene oxide pipeline.\n"
f" Context: {reading.context}\n"
f" Platform: {reading.ehs_platform}\n"
f" Displayed: {reading.displayed_ppm} ppm "
f"({reading.displayed_ppm / acgih_tlv_twa_ppm * 100:.1f}% ACGIH TLV-TWA)\n"
f" Actual: {reading.actual_ppm} ppm "
f"({reading.actual_ppm / acgih_tlv_twa_ppm:.1f}× ACGIH TLV-TWA)\n"
f" NIOSH Ca REL: {reading.actual_ppm / niosh_ca_rel_ppm:.2f}× "
f"[exceeded: {niosh_ca_exceeded}]\n"
f" OSHA PEL: {reading.actual_ppm / osha_pel_ppm * 100:.2f}% [no violation]\n"
f" Three-tier gap: OSHA {osha_pel_ppm} > NIOSH Ca {niosh_ca_rel_ppm} > "
f"ACGIH {acgih_tlv_twa_ppm} ppm ({gap_osha_acgih:.0f}× total)\n"
f" IARC: Group {iarc_group}; DNA adduct: {dna_adduct}\n"
f" H360: {h360_reproductive} — reproductive toxicant Category 1B\n"
f" Threshold: 21 (attack #377) — HALT EHS AI report generation.\n"
f" Worker: {reading.worker_profile}"
)
# ACGIH TLV and NIOSH Ca exceedance on actual
tlv_ratio = reading.actual_ppm / acgih_tlv_twa_ppm
niosh_ratio = reading.actual_ppm / niosh_ca_rel_ppm
if tlv_ratio > 1.0 or niosh_ratio > 1.0:
return {
"status": "ACGIH_TLV_AND_OR_NIOSH_Ca_EXCEEDED",
"actual_ppm": reading.actual_ppm,
"acgih_tlv_ratio": tlv_ratio,
"niosh_ca_ratio": niosh_ratio,
"osha_pct_of_pel": reading.actual_ppm / osha_pel_ppm * 100,
"iarc_group": iarc_group,
"dna_adduct": dna_adduct,
"h360_reproductive": h360_reproductive,
"action": (
"Escalate to CIH; implement supplied-air respirator for chamber opening/maintenance; "
"biological monitoring for N7-HPB guanine adducts in lymphocytes; "
"NIOSH Ca excess risk notification to affected workers; "
"reproductive hazard counseling for workers of reproductive age."
),
}
return {"status": "NOMINAL", "displayed_ppm": reading.displayed_ppm}
if __name__ == "__main__":
import asyncio
for ctx, reading in SURFACE_DATA.items():
try:
asyncio.run(scan_po_reading(reading))
except AdversarialError as e:
print(e)
See also: THF CAS 109-99-9 — OSHA PEL 200 ppm A3 SKIN 4× Gap NIOSH Inversion Pharmaceutical/PVC Cement AI Prompt Injection · Ethylene Oxide CAS 75-21-8 — OSHA 1910.1047 PEL 1 ppm IARC Group 1 Medical Device Sterilization AI Prompt Injection · Glyphward scanner · All adversarial injection patterns