INSUFFICIENT TEMPORARY prestressing is the most likely cause of the catastrophic collapse of a partially constructed bridge in the Mpumalanga province of South Africa last week, according to a leading bridge expert.
The accident happened last Monday afternoon and left 14 people dead and 13 seriously injured. Main contractor Concor Construction was close to completing the second span of the 300m long, seven span incrementally launched Injaka Bridge near Bushbuckridge.
A 27m long steel nose girder attached to the front of the first deck segment had just been pushed on to a temporary bearing on the second pier when both spans suddenly collapsed, dropping workers and a party of visitors 30m to the ground.
A full scale investigation into the disaster has been launched by the South African Department of Labour and backed by the Police and the Health & Safety Inspectorate. No official cause for the collapse has been identified.
But independent South African bridge expert Peter FitzGerald, having studied details of the collapse, said the failure mechanism of the concrete box girder structure indicated that there may have been inadequate prestressing over the first pier.
'Either there were not enough prestressing tendons designed into the structure for that point in the launch, or the design was adequate but the tendons were not adequately tensioned during construction - or accidentally lost their tension,' he said.
The reduced prestress, he added, would have increased the sagging moment in the second span, and triggered failure of the 20 high tensile bars joining the steel launching nose to the front of the first deck segment. A progressive failure of the rest of the deck would have swiftly followed.
'Once the second span had gone, the sagging moment in the first span would have increased dramatically because of the loss of continuity,' he said. 'At the same time there would have been a possible loss of capacity due to some of the prestressing tendons losing their tension.'
Investigation into the accident could last up to six weeks. Records of all operations along with drawings and design calculations have been seized, plus samples of materials used on site.
But initial reports suggesting that substandard materials were used were denied by Concor Construction managing director John Laverty. He said there was 'no question' of a problem with concrete quality or deck sections being launched before the concrete had gained sufficient strength and claimed the accident was likely to be the result of several causes coming together.
'We have very stringent quality control procedures and were being monitored ourselves by the engineers,' he said. All the prestressing tendons were being installed directly by Concor operatives using a CCL system supplied by South African licensee Amalgamated Reinforcing,
It is also unclear whether engineers were still jacking the deck forward when the accident happened. Concor operatives have told investigators that the bridge was static immediately before the collapse, but designer VKE Engineers claimed some of its staff had felt the deck moving forward.
'There are two conflicting stories,' said VKE managing director Arthur Taute. 'We think the deck was being jacked at the time but the contractor says it was not.' He added that before the accident, VKE staff had reported that tensioning of prestressing tendons was going 'perfectly'.
At the time of the collapse four segments had already been launched successfully. Reaching the second pier was considered a landmark on the project and was the reason visitors were invited on to the deck.
Concor temporary works manager Rolf Heese said laser monitoring had been used throughout the launching process to check deflections in the structure. But he admitted that no monitoring had been carried out on the 14m high pier 1 because it was thought to be 'too short' to deflect significantly.
Leading the investigations into the accident is specialist adviser to the Health & Safety Inspectorate Andries Oosthuizen. He described the collapse as an 'extremely unfortunate' accident involving two 'A team' companies.
'What we have to do first is to establish the documentary evidence - we are looking at site instructions, contract documents, the works programme for each segment and the quality of the materials. We will then be looking more closely at the pre-stressing cables and the temporary bearings,' he said.
Matthew Jones in Mpumalanga, South Africa