Last night at Champtoc^-sur-Loire, near Angers, a car went over the parapet of a bridge over the Paris-Nantes line and was struck by a
TGV▸ . The driver was killed, and the train took some 2 km to stop. The passengers were evacuated to another TGV, which took several hours, and one carriage had to be lifted by crane to examine the debris of the car before the track was cleared.
This line is not custom-built
LGV▸ , indeed there is a little station at Champtoc^-sur-Loire. The train was reported travelling at 200 km/hr, so either the line has been uprated or the reporting is not based on line speed. The road (D15) is pretty straight, with a light handrail and an ARMCO barrier in front of it and extending to part of the approach roads. However, there are unprotected parts of the approach, and the barrier has the traditional downward slope at the ends of the bridge. But on the whole the protection of this bridge is not obviously out of date or deficient, so this may not lead to a big panic programme in France.
There is not much information yet on how the car got onto the track, but one picture shows the front skirt of the body left on the bridge behind the ARMCO (
http://www.ouest-france.fr/actu/actuDet_-TGV-Paris-Nantes.-Le-train-percute-une-voiture-tombee-d-un-pont-un-mort-_40774-2237775_actu.Htm). That suggests that the car did go over the parapet, rather than to one side of it. This possibility, of the slope at the end of the barrier forming a ramp, was not one that featured in the review and upgrading process after Great Heck in Britain - at least it is not in the
DfT» guide on the subject (
http://assets.dft.gov.uk/publications/tal-6-03/the-guidance-booklet.pdf). ("Ramped end terminals" have generally gone out of favour, and or not now allowed by DtT on roads above 50 mi/hr.)