Near misses
The near miss nobody heard about: TWA 514, December 1974
Six weeks before TWA 514 hit a Virginia mountain, another crew made the same mistake and got away with it. Nobody else was told. Then aviation changed.
The Ghost Patch opens with an epigraph from a fictional insurer’s handbook: “A near miss is a claim that has not yet been paid.” The handbook is invented. The idea behind the line is real, and one accident in 1974 turned it into a working system.
1 December 1974
TWA Flight 514 was coming into Washington Dulles in cloud and turbulence. The crew misread an air traffic control clearance and descended early. The aircraft hit a mountaintop in Virginia, and everyone on board died.
That’s the accident. The part that matters here happened six weeks before it.
The crew who got away with it
A United Airlines crew had made the same misunderstanding of the same clearance, and they narrowly missed the same mountaintop. After landing, they reported it, and United sent its own pilots a cautionary notice.
Then nothing more could happen, because nothing existed for it to happen through. NASA’s own history of the programme says there was no method of sharing the United pilots’ knowledge with TWA or any other airline, so a warning that was real and written down stayed inside one company.
Nobody hid anything. That’s what makes it uncomfortable.
Who should hold the reports
The FAA wanted a way to collect incident reports from across the industry and pass the lessons on. The idea that came out of it was a national reporting programme that would be voluntary and confidential, with no punishment for the person reporting.
The FAA didn’t run it. An agreement in August 1975 gave NASA the job of setting policy and running day-to-day operations, while the FAA funded and sponsored the programme and supplied its immunity provisions, which meant the regulator paid for a system whose individual reports it was not allowed to ask for. NASA was a neutral third party. The Aviation Safety Reporting System began operating in April 1976, sixteen months after the crash.
What the reporter gets
The terms are strict, and they’re the reason it works. The FAA won’t seek the reports, and NASA won’t hand them over. The FAA doesn’t use them in enforcement, except for criminal offences and accidents. Every report carries an identification strip with the reporter’s details, and NASA time-stamps it, returns it to the reporter as a receipt and keeps no copy, accident and criminal cases aside, so the main thing that could tie a name to the report goes back to the person who filed it.
What’s left is an account of what went wrong with nobody’s name on it.
One reading is that the choice of who held the reports mattered more than anything else about the design. A pilot who has just come close to a mountain has good reasons to tell the employer and the regulator as little as possible. Tell a neutral body that can’t punish you, and the reasons weaken. NASA’s own account says only that a neutral party was chosen to run a programme built to be voluntary and confidential. The rest is inference.
Where The Ghost Patch comes in
The novel’s blurb has a process that refuses, with forty-one seconds left on the clock, to let a machine switch off a dialysis feeder. By the aviation definition that is a near miss. It also belongs to a machine that the company owning it has a lawful reason to erase by Friday.
Real organisations meet smaller versions of that all the time. A job fails and is quietly rerun. A disk fills and someone clears it at midnight. A certificate lapses, and a colleague spots it with an hour to spare. Each one is a near miss, and most are never written down, because nothing went wrong.
If you run systems, ask one question this week. When something nearly goes wrong at your place, who hears about it? If the answer is “whoever was on shift”, you’re on United’s side of 1974.
The Prologue and Chapter 1 of The Ghost Patch are free to read on this site.