The calendar for 2026 is the same as that of 1987. Which means that Sunday, January 4, is exactly the same day of the week as the one in 1987, which brought us the tragedy of the Chase, MD Amtrak wreck of Train # 94, the Newport News, VA to Boston Colonial. with three Conrail light engines at Gunpowder interlocking on the Northeast Corridor. The cause was the intoxication and negligence of the Conrail crew, engineer Ricky Gates and fireman Ed Cromwell, who had beer (possibly pot, and possibly a portable TV with them to watch football playoffs); and had overridden or defeated more than one federally mandated safety device that might have alerted them to the stop signal at the end of the main track they were on at Gunpow interlocking, at the west end of the bridge over the mouth of the Gunpowder River, where the NEC goes from four main tracks to two across the bridge. Realizing far too late that they had a stop signal at Gunpow, they slid to a stop with two engines past the switch, fouling the interlocking. Meanwhile, Amtrak Train #94, with AEM7s #903 and 900 and 18 Amfleet I coaches, nearly full of end of the holiday period travelers, and running nearly an hour late out of Baltimore, was just behind them trying to make up time, bending the non-Metroliner speed limit of 110 mph to about 128 mph.

The Conrail engines slid past the stop signal seconds after #94 passed the last clear signal before Gunpow, so as far as engineer Jerry Evans knew he had a clear track ahead and did not see the fouling locomotives until he came around the curve before the bridge, about half-a-mile away. All he had time to do was throw the train down into emergency and make one last radio call to the conductors on the train to brace for a collision, only slowing the long, heavy train by about 3 mph before impact. Evans was beheaded in the wreck, one of 16 passengers and crew who died that day in what was the worst wreck in Amtrak's history up to that point. If the train crew of 94 had not been holding a deadhead cafe first car empty for oncoming passengers in Philadelphia, the death toll could easily have exceeded 50, because that car was flattened in the wreck.

It took three days for Amtrak to clear enough of the wreckage to get one track reopened. I know that because I was LSA on the lounge of one of the first trains go by the wreck site, the northbound Montrealer. Traveling to Wilmington with us that evening was Washington Division Manager Stan Bagley, who was on his way to meet with Jerry Evans' widow. I had arrived in DC three days earlier on the southbound Montrealer, on #25 track in Union Station, right beside Train #94 on #26 track (the longest lower level through track), which was waiting for us to clear before they departed for disaster after adding about 10 additional cars to what had come up from Virginia. I watched it leave on it's short journey to infamy from the lounge as we waited for the commissary crew to take our remaining stock away. I heard about the wreck a couple of hours later as I was cashing-out the end of my trip at the remittance office by track #7, and spent the next three days watching news coverage of the wreck and reading Don Phillips' extraordinary reporting on it in the Washington Post. It almost wasn't #94 that was in the wreck. Because it was running so late, the dispatcher held an extra holiday Metroliner in Baltimore for it to get ahead of them to try and make-up time. I know that because a co-worker friend was working the club car on that train, and the first cars of that much shorter Metroliner were all full, so the death toll if that had been the train to hit the Conrail engines would certainly also have been much higher.

This anniversary is relevant to this site for a couple of reasons. First, the reverberations from it significantly changed the rules which govern not only regular railroad operations, but also heritage and excursion operations that fall under FRA jurisdiction. Mandatory random drug and alcohol testing of train crews and federal licensing of engineers came out of that accident, as did tougher rules on the protection and use of mandatory locomotive safety appliances. And an argument could be made that this wreck made it easier for Congress to pass the 2008 Positive Train Control rules after the CalTrain wreck of that year, which has added to the obstacles for operating heritage equipment (like Bennett Levin's E-units). Less significantly, we lost the first AEM7 - #900 - in that wreck, which might have ended up in a museum.

So, at the end of this holiday period and the start of a new year, let us pause to remember that terrible day that brought many changes to the industry and avocation many of us love and care about.

I've always remembered the wreck for the following reason. My dad worked between Washington and New York (and other places when he had less seniority) from 1969 until 1999. At the time of the wreck he was the regular conductor for 94. He preferred to work
the Florida trains when he could (Silver Meteor and Silver Star) but again 94 was his regular job at that time. That particular day he marked off and another conductor took his place.
Dad knew both the engineer and the other conductor well. He never talked much about the
wreck except that he had heard the conductor while surviving the wreck had one of his ears cut off and that the wreck was so bad that one of the engines (I don't know which one) had been stripped down to the frame in the wreck. He would complain about the Norac rules that came out of this. If he hadn't taken off that day he would have been in it. I have no idea if it ever bothered him about what happened and that he hadn't been on it that day. He never talked about it.

903 was the lead unit that was stripped down to the frame, which ended up in the river. I can tell you that even though she was not directly involved in the wreck, my co-worker friend who was on the Metroliner that 94 passed had survivor guilt issues for years afterwards.

I wasn’t there, but remember Rob Hoffer talking about it on TrainNet back when. The Silver Star was one of his usual assignments. He lived fairly close by. In fact, he was just home from a run when he heard this one nearly out his back door and ran out to help:

https://www.ntsb.gov/investigations/Acc ... AR9702.pdf

Corky Price wasn’t on duty the day of the Gunpowder wreck, but he had to look at the wreckage because of his place in the BLE. He hadn’t gone to VRE yet and I wonder how much that wreck, along with several other horrible incidents, had to do with his move.

ETA: while TrainNet didn’t come along till 1989, events like Gunpow cast long shadows. As the old saying goes, the rulebook was written in blood.

I knew Rob. We used to compete on the Capitol Ltd. to see who could give the best narration between DC and Cumberland on the historic points of interest over the PA.

Man, that would have been some ride. Is Rob still with is? Corky (Corbett Price Sr.) had heart issues and died a little while after he retired.

The DC area seems to have a disproportionate number of wrecks, but traffic is so heavy and so many different railroads operate (VRE, Amtrak, DC Metro, MARC…) that the law of averages bites even without substances and distractions .

And since this is now 2026, and social media and AI are contaminating everything, I noticed a 'photo' of the accident under a 'story' of the anniversary of the accident on FB that was a AI generated piece of garbage showing an AEM7 turned over sideways in a cut after a head-on collision with a three unit CSX freight train. Photorealistic.

The level to which "historic" images in particular are contaminating historical fact is alarming. No need to find the right picture if you can have AI generate something for free.

Agree. AI exacerbates the problems that Photoshop and the like have caused researchers.

This may seem unrelated, but it’s not: there is a group of Civil War reenactors who pride themselves on being as accurate as humanly possible. They even have a wet-plate photographer, and there lies the difficulty: someone laid hands on a few copies of reenactor photos and marketed them for premium prices as period originals. Unless you recognize the people in the photo, you’d be very hard put to know which picture was real 1860s vintage. The same is true of railroad pictures taken in black and white, with crew in clothing correct for the year.

Now consider the wrecks. If what a future researcher has to look at is a skillful AI image, it may lead to wrong conclusions and even rulemaking. Provenance will be key. Since railroading often strains credibility at the best of times—the OR&W managing to wreck four trains in one day, for instance—bored people or bad actors may easily alter the details of accidents, even in videos, or invent wrecks where there were none.

AFIK Rob is still around, haven't heard anything to the contrary.