Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Viaduct parapets probed in fatal Italian coach crash

Inadequately tied in precast concrete parapets could have allowed an out of control coach to plunge 30m from an Italian motorway bridge, killing 38 people, senior engineers told NCE this week.

The crash happened on the A16 motorway running west-east between Naples and Bari, near the town of Avellino in the Campania region of southern Italy. The coach was travelling west towards Naples and left the viaduct on the inside of a sharp curve, 30m above ground level.

Avellino coach crash location

Avellino coach crash

The coach driver is believed to have lost control of the vehicle which crashed through the parapet on the Acqualonga Viaduct (see map).

The coach hit a number of cars before ploughing through what Italian authorities described as a New Jersey type barrier – a parapet made up of standard precast concrete units of the type seen in use commonly around the world (see box).

Images of the barrier after the crash, some of it lying in the ravine 30m below, and of the viaduct deck itself suggest the units were not tied in to the deck, engineers said.

A statement from Autostrade per l’Italia, the concession company which owns and operates Italy’s motorways, said the New Jersey parapet had been fitted with guardrails and was specified only to absorb the shock of colliding passenger cars, which make up the vast majority of impacts.

Shea Concrete Products of New Hampshire in the United States is a major supplier of New Jersey type barriers. Its product manager Greg Stratis, a member of the National Precast Concrete Association board, told NCE that any permanent structure of this nature should be tied in or thoroughly embedded in the bridge slab.

“For a temporary application you may just sit it down [on the deck or road surface] but here in the US we would still bolt it down, maybe using epoxy resins,” he said. “But if it is a permanent structure the barrier would be embedded nine to 12 inches [229mm to 305mm] and if there was any chance a vehicle could fall off a bridge or hit oncoming vehicles, yes, you would tie it in.”

Forensic engineer Rajen Lavingia of Glasgow-based consultant Cadogans added: “Photographs in the public domain that I have seen appear to show that there was no positive connection between the barrier and the bridge deck. “It is my opinion that it was a plain mistake. Someone has either forgotten to attach the barriers to the slab or the connection(s) between the barriers and the slab was (were) poorly designed and executed.

“If the barrier had been tied in with reinforcement bar it would have been probably okay. This implies that accident would not have occurred.

“As a structural engineer I would have ensured that the barrier was tied in.”

Jersey barrier history

Jersey barrier

Source: US Federal Highways Administration

Typical F-shape concrete barrier section

Although now used as a generic term to describe a range of precast concrete road barriers, the original Jersey barrier was invented in the 1950s by the Stevens Institute of Technology, New Jersey as a means to temporarily divide highways.

Although these barriers can be slip formed insitu, they are predominately made of precast reinforced concrete.

The units have a gently sloped ramp at the base designed to slow a crashing vehicle while absorbing impact with controlled deflection. There are many variants of the original Jersey barrier (including the F-type shown).

In the UK a variant called concrete step barriers are used. This drawing shows that such a parapet design incorporates steel reinforcement to tie the barrier into the edge of the deck.

Lavingia was keen to stress that not all accidents could be designed out. “It is important to note that even with a properly designed and constructed barrier, the coach might have gone through,” he said.

Atkins highways and transportation head of technology Peter Whitfield described the apparent lack of tied in connections as “unusual”.

“I would be amazed if it complied with the requirements,” he said. “It would be normal for a parapet like this to be tied into the bridge deck. The investigators will be looking to see if the tie in detail for the parapet and the bridge met with the required design standards at the time.”


Google Street View images suggest that the parapet is a bolt down type secured by up to four bolts. The images also indicated that in certain sections these bolts were missing, as were lateral connections between the units which are supposed to hold the precast sections together.

The level of containment required of a barrier system is determined by risk analysis. Lavingia said that the gradient of the road, the nature of the bend and the height of the Acqualonga viaduct meant that engineers should have viewed it as a medium to high risk structure.

“Risk is calculated by multiplying probability by severity,” he said. “The probability of a plane crashing is low but the severity is very high; combined, the risk is also low.

“However, the likelihood of this sort of accident, where you [appear to] have a coach out of control is definitely not a low probability, and of course the severity is extremely high – combined, the risk would have been in the medium to high range.”

News agency Reuters reported that Italian magistrates were investigating the role of Autostrade per l’Italia in the accident. At the time of going to press the firm had not responded to NCE’s requests for information.

Parapet design

Consultant Atkins highways and transportation head of technology Peter Whitfield says vehicle weight, speed and angle of impact influence the behaviour of a bridge parapet. This also depends on the design criteria used, in particular the specified containment level.

A key issue for designers is the selection of the level of containment.

In lower risk scenarios a “normal” level of containment is specified. This would be sufficient to contain a 1.5t car at 110km/h.

A “higher” level of containment would withstand a 13t coach at 70km/h and the highest specified containment level, known as “very high”, would be used for high risk scenarios such as where a railway runs below a bridge. This would withstand the impact of a 30t HGV travelling at 65km/h.

All vehicle containment parapets must comply with the Test Acceptance Criteria requirements of BS EN 1317-2 based on the above vehicle weights, speed and also an impact angle of 20°. Higher impact angles can occur if a vehicle has rebounded or has struck another vehicle.

The UK design standard is TD 19/06 (Design manual for roads & bridges, volume 2, section 2, part 8) and this requires the risk level to be assessed before deciding what level of containment is specified.

In the UK a long viaduct which does not cross a railway would normally be provided with a parapet that would withstand, as a minimum, a 13t coach impact.

UK standards were reviewed after the Selby rail crash in 2001 when a vehicle crashed off the M62 onto the East Coast Mainline. While no major shortcomings were identified with the standard, the risk theory based Road Risk Restraint Standard was produced as a means to address the industry’s concerns about how risks were previously assessed on the UK’s highways.

Readers' comments (1)

  • Incompetence? Corruption? Who knows? It will be interesting to see if a full enquiry really is carried out and made public.

    Unsuitable or offensive? Report this comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.