The Records Said Seamless

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

A neighborhood in San Bruno blew up on a Thursday evening in 2010, and the pipe that did it had a defect the operator’s own records swore wasn’t there. The paperwork said seamless. The steel said otherwise. Here’s the day it happened, and the lesson every operator who trusts a drawing needs to sit with.

At 6:11 p.m., a 30-inch natural-gas transmission pipeline — PG&E’s Line 132 — ruptured under a residential street in San Bruno, just south of San Francisco. Escaping gas found an ignition source and went up. The blast left a crater in the road, destroyed 38 homes, damaged dozens more, injured 58 people, and killed eight.

Then the part that turns a rupture into an indictment. It took roughly an hour and a half to get the gas shut off, because the valves had to be closed by hand by crews driving to the site — there was no automatic or remote way to choke it off while a neighborhood burned.

And when the National Transportation Safety Board pulled the failed section (report NTSB/PAR-11/01), they found the rupture started at a seam weld in a section of pipe installed in 1956 — a weld so poorly made it never should have passed. PG&E’s records described that stretch of Line 132 as seamless pipe. It wasn’t. The company had been operating and pressure-testing a line for decades on paperwork that did not match the steel in the ground, and its integrity-management program never identified the seam as a threat, because on paper the threat didn’t exist.

Nobody at the sharp end on September 9 did the thing that killed those people. The lethal decisions were made in records rooms and engineering offices over fifty years: a bad weld waved through in 1956, a records system that carried the error forward as fact, an integrity program that trusted the records instead of the pipe, and an emergency setup that couldn’t stop the gas fast enough once it let go.

The HOP read: this is what it looks like when your map is wrong. Every downstream decision — what to inspect, how high to pressure it, whether to worry about seam welds — was rational given the records, and every one was built on a false foundation. You cannot inspect your way out of a bad record, because the record tells you where not to look. The barrier that failed here wasn’t a valve or a weld. It was the assumption that the paperwork was the reality.

Gas transmission pipelines sit under PHMSA at 49 CFR Part 192 (current as of September 9, 2026 — verify at ecfr.gov). Integrity management for transmission lines lives at 192.917 and 192.921 — you have to identify the threats a segment actually faces, and manufacturing and construction defects like a bad seam weld are on that list. Maximum allowable operating pressure is governed by 192.619. San Bruno is the case that drove the requirement that operators show records that are traceable, verifiable, and complete — not “the file says seamless.” It also fed the federal push, in the 2011 pipeline safety law, toward automatic and remote-control shutoff valves so a line can be isolated in minutes, not while a block burns.

You may not run a pipeline. But the floor-level version is 29 CFR 1910.119(j) — mechanical integrity for covered processes: your P&IDs, your relief-valve set points, your material-of-construction records are only as good as their accuracy, and 1910.119(d) requires that process safety information be complete and correct. The San Bruno question for your site is simple: if a drawing is wrong, how would you ever know — and what decision are you making right now because a record told you a threat “isn’t there”?

Most incidents are E, H, and S at once, and this one hit all three by design — a transmission main runs through where people live.

Environmental: a ruptured gas main is an uncontrolled release before it’s a fire — methane straight to atmosphere at line pressure — and once it ignited, the products of a 38-home fire went into the air and the runoff and ash into the storm system. The pipe, the flange, the tank shell is your containment, and a containment failure is a reportable environmental event before it is anything else. Asset integrity is environmental compliance wearing steel.

Health: eight dead, 58 injured, most of the acute harm from burns and blast. The slower health story is the neighborhood that watched it and the responders who worked it — trauma is an exposure, and it doesn’t clear when the fire’s out.

Safety: covered above — a wrong record is a hazard you can’t see, because it’s the thing telling you where the hazards aren’t.

The federal rule meant to protect fenceline communities is being loosened — watch it. EPA’s proposed RMP “reconsideration” would roll back major pieces of the 2024 Safer Communities by Chemical Accident Prevention rule — the safeguards for facilities holding hazardous chemicals near neighborhoods like the ones around San Bruno’s pipeline right-of-way. A final rule is pending. So what for safety leaders: if you run a PSM- or RMP-covered process, some requirements you’ve planned around — third-party audits, root-cause after incidents, sharing hazard information with local responders — may change. Track the final rule, but don’t quietly delete good practice because the legal floor moved. Your fenceline neighbors don’t get a comment period on the next release. (Source: EPA RMP, epa.gov/rmp.)

Shared blamelessly, because that’s the only way anyone learns from it.

A maintenance planner is scoping a hot-work job on a “spare” process line the P&ID shows as out of service and blinded. Standard prep, permit in hand. A pipefitter walks the actual line before the torch comes out — physically, hand on the pipe — and finds it isn’t blinded where the drawing says; a blind was pulled during a turnaround two years ago and the drawing was never redlined. The line still had product to a valve twenty feet away. They stop, verify, blind it for real, and only then light up. Nobody gets hurt, because somebody trusted the pipe over the paper.

The HOP read: the planner did everything right against the information he had — and the information was wrong, carried forward as fact the same way “seamless” was carried forward on Line 132. The catch wasn’t a better procedure. It was a person close enough to the work to check the steel against the drawing, at a site where walking the line before hot work is expected and thanked. The fix isn’t “read the P&ID more carefully.” It’s “the drawing is a hypothesis until someone confirms it at the pipe.”

Got a fail or a near-miss? Hit reply, or use the link below — we’ll feature it anonymously, no names, no company, no blame. Just the lesson. Share yours here → https://scantoconfess.com

Pick one drawing your people are betting on today — a P&ID, an isolation list, a pipe spec, an as-built — and go confirm one thing on it at the physical asset. Is that valve really where the drawing says? Is that line really blinded, that material really what the spec claims, that record really matching the steel? If it matches, good. If it doesn’t, you just found the gap between the map and the ground — and that gap is exactly what turned a records error from 1956 into eight funerals in 2010.

🎬 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 📤 Forward this to a crew member who needs it today.

Please stay Safe & Hydrated!!!