Leak Wouldn’t Freeze

Those troubleshooting the problem originally assumed the ammonia would freeze inside the car and at the valve as the pressure inside the car went down. If the valve had frozen for even a minute, a crew could have replaced it and stopped the leak, on-site personnel said.

Local fire departments covered the tanker with a fine spray of water, trying to keep the ammonia gas fumes contained. That water froze despite the 80°F-plus ambient temperatures.

Why the valve and ammonia didn’t freeze could not be explained. Experts had hoped the freezing would occur at about a 5-lb/sq-in. pressure. That level was reached close to 48 hrs after the leak was first reported.

The next effort involved using nitrogen, an inert, nontoxic, nonexplosive gas. The nitrogen, under pressure, was pumped into the tank car, forcing the liquid ammonia out of the bottom, through pipes, and into a second rail car. The nitrogen filled the top of the car as it was pumped in and bled through the leaking valve, replacing the ammonia that was leaking.

EPA officials on the scene said all readings other than those immediately around the rail car showed levels of ammonia well within the amounts that the government holds to be safe, even with prolonged exposure. Although some 30 to 50 people were initially taken to area hospitals, mostly with breathing complaints, officials reported that none were serious and only the most fragile were held, largely for observation.

A task force of federal, state, and Canadian officials (a Canadian rail carrier transported the ammonia tank car) investigated the cause of the leak, and have released some of their preliminary findings. The Federal Railroad Administration said it is focusing at least part of its investigation on the Trinity Saginaw Facility in Saginaw, TX, where the tanker car last underwent repairs in July. One source close to the investigation said it was the tanker’s first trip since those repairs.

Publication date: 10/02/2000