The Crack Was Already There

▶️ Prefer to watch? Every story’s a short video → https://www.youtube.com/@GetSafetyIntel · 📣 Got a near-miss? Share it anonymously (QR + link at the bottom).

Some failures happen in a second. This one took ten years. A crack that started in a repair weld back in 1974 grew quietly — inside the steel, where nobody could see it — until a worker spotted it on a routine walk-around and there was nothing left to hold it together. Today we’re on hydrogen sulfide and pressure vessels, and the same invisible gas runs through two of our three stories.

No verified major workplace-safety incident lands on exactly July 20, so we’re pulling from this same week — and this one earns your five minutes.

Shortly before 6 p.m. on July 23, 1984, a worker at the Union Oil refinery in Romeoville, Illinois — just southwest of Chicago — was making his routine hourly check of an amine absorber tower. That 62-foot tower had one job: strip hydrogen sulfide out of a propane-and-butane stream. On his walk, he found a hairline crack in one of the tower’s welds. He moved to close the inlet valve and stop the leak. The crack didn’t wait — it ran from about six inches to two feet, and a cloud of flammable gas poured out. The plant’s own fire brigade was moving in when the cloud found an ignition source.

Seventeen men died. Seventeen more went to the hospital. The blast tore the top 46 feet off the tower and threw it two-thirds of a mile. More than $100 million, gone.

Here’s the part that should stick with you. When the National Bureau of Standards cut that steel apart, the tower hadn’t failed because it was overloaded. It let go at roughly 10% of the stress that steel was rated to hold. Back in 1974 — ten years earlier — the refinery had welded in a replacement section of the shell and skipped the post-weld heat treatment, the step that relieves the stress welding leaves behind. That hard, brittle weld sat in a wet hydrogen-sulfide environment and cracked from the inside out, hydrogen working its way into the metal until the steel had lost more than half its toughness. By the day it failed, a crack had already eaten through more than 90% of the wall. The worker didn’t find a new problem. He found a ten-year-old one, on its last day.

OSHA answered this one directly. On April 11, 1986, it sent a formal memo to every regional office — a real document you can still pull up today — warning that amine absorber vessels across the industry were cracking, and that field repairs welded without proper heat treatment were setting up brittle failures. A follow-up survey found roughly 60% of the amine absorbers checked had cracks. This is the DNA of the Process Safety Management standard, 29 CFR 1910.119 — and specifically mechanical integrity, 1910.119(j): you inspect and test your pressure vessels on a schedule, you fix the deficiencies before you run again, and you make your repairs to code.

If you run pressure equipment, “it looks fine” is not an inspection. The crack that kills you is on the inside of the wall, and it may have been there longer than the new guy’s been alive. Mechanical integrity means somebody with the right tools — ultrasonics, magnetic particle — actually looks, on a schedule, and a bad repair from two shifts ago (or two decades ago) still counts against you today.

(E) Environmental — what let go was flammable hydrocarbon gas from a unit whose whole purpose was to trap hydrogen sulfide, so when it burned, it burned as a refinery fire: hydrocarbons and sulfur compounds going up as black smoke over the Chicago suburbs, plus a 46-foot section of steel thrown two-thirds of a mile as shrapnel. The documented damage was on-site and enormous ($100M+); I don’t have a verified record of offsite water or soil contamination, so I won’t invent one — here the environmental hit was the fire and what it put in the air. (H) Health — the acute toll is clear and brutal: 17 dead, 17 hospitalized, mostly burns. And the gas that tower existed to remove, hydrogen sulfide, is a killer in its own right at high concentration. (Tie it together) One cracked weld was a safety failure, an air-quality event, and a standing reminder that the H2S you’re managing is a health hazard the entire time — three faces of the same bad weld.

It’s July, it’s hot, and here’s what matters. OSHA still has no final heat standard — the rule it proposed in August 2024 is stuck, the hearings wrapped in 2025, and there’s no finish date on the calendar. But on April 10, 2026, OSHA rolled out an expanded Heat National Emphasis Program that runs five years. Translation: a compliance officer can show up and inspect on any day the heat index is forecast to hit 80°F, no employee complaint required — and if you’re not protecting people, they cite you under the General Duty Clause, Section 5(a)(1) of the OSH Act. “There’s no heat rule yet” is not a defense. Water, rest, shade, and easing new workers in over their first week — that’s the whole game.

On July 14, the Chemical Safety Board released an update on the hydrogen-sulfide release that killed two young workers at the Woodland Pulp mill in Baileyville, Maine, back on January 27. A 20-year-old co-op student and a 26-year-old engineer — both new — were on a second-floor drawing project, unrelated to the plant shutdown happening around them. During that shutdown, high-pH, sulfur-bearing fluid drained into a long “acid sewer,” an automatic system kept dosing it with sulfuric acid, and the two reacted to make H2S. Then someone shut off the scrubber fan as part of the shutdown, and the gas had nowhere to go but into the building. CSB’s findings so far: no personal H2S monitors, no fixed H2S detectors in that area, no building ventilation, and no system tracking who was even in the building. The two weren’t found for hours. So what for you: startups and shutdowns rewrite your hazard picture in real time — the moment you turn a scrubber off, the gas math changes — and a personal monitor is the difference between a bad day and a funeral.

A crew’s about to open a sump that’s thrown H2S before. The tech grabs his four-gas monitor out of the truck, powers it on, sees it light up green, clips it on. A newer guy asks, “when’d you last bump-test that?” Nobody knew. They walked it to the test station and hit it with cal gas. The H2S sensor never alarmed. Dead sensor, cheerful green light. If they’d trusted the green light, the first warning they’d have gotten was on the ground.

Nobody screwed up here. The monitor powered on and looked healthy — the design lets a dead sensor look exactly like a good one, and “bump test” lives on a sticker nobody reads. The fix isn’t “be more careful.” It’s a bump-test station at the gate you physically walk past, and a rule that a monitor without a bump test today counts as no monitor at all. The newer guy asking the dumb-sounding question is the whole system working.

Got a fail or a near-miss? Hit reply, or scan the QR below — we’ll feature it anonymously — no names, no company, no blame. Just the lesson, so the next crew doesn’t learn it the hard way.

Before your next shift anywhere hydrogen sulfide can show up — a sump, a tank, an acid sewer, a sour-gas line — bump-test your personal monitor and watch it actually alarm. Not power on. Alarm. And if there’s no monitor for a spot where H2S can form, that’s your one conversation with your supervisor today.

🎬 Rather watch than read? Every Safety Intel story is a short video on YouTube → https://www.youtube.com/@GetSafetyIntel — subscribe so you never miss one. 📸 Share YOUR near-miss (100% anonymous): https://scantoconfess.com — or just scan the QR above. 📤 Forward this to a crew member who needs it today.

Please stay Safe & Hydrated!!!