selection biases and risk assessment

flicking across cnn.com, I see a propane train derailment and an oil well fire in Texas

is it heigtened awareness, broader reporting to fill cable news shows, or is there a higher incidence of petrochemical and industrial chemical accidents this year?

The EPA RMP*Info list of industrial accidents and sites is apparently confidential, although it was originally compiled for public information, but there is a summary of five year incidence in the late 90s in a public report.

Glancing through the data, there seem to be refinery/extraction accidents about once per week and transport/LNG accidents about once per month - now there have been several "recently", like the burst oil pipeline in Texas and a couple of spectacular chemical plant explosions this winter, but memory compresses events like that and it seems impossible to get an actual statistic on the rate of such events.
Not to mention that the "events" are hard to properly define as a prior - I suspect "telegenic" is not an EPA incident category.

Someone out there have actual access to the data?

Tags

More like this

When cars are stocked with airbags in every possible direction - are there ceiling airbags yet? - drivers become more aggressive:
If you just want to look at accidental death, I would note that most of the decrease in fatal gun accidents in the US occured before there was an increase in handgun ownership
Accidents in swimming pools can be serious or fatal (drowning, broken necks) but fortunately they are rare. For pool operators one of the more likely nightmares of daily operation is when someone has "a fecal accident." In other words, someone craps in the swimming pool.
Although alcohol consumption plays a role in about 31 percent of homicides, only 1.4 percent of TV news reports on murders mention alcohol. Only 12.8 percent of TV news stories on traffic accidents mention alcohol, while 34 percent of accidents involve drunk drivers.

"The EPA RMP*Info list of industrial accidents and sites is apparently confidential ..."

We wouldn't want to embarrass any of our industry contributors, now would we?