Back in 1995 it seemed more than reasonable to save the £750,000 cost of a detailed condition survey of the 20 year old steel box girder Avonmouth bridge. The structure had been continuously monitored and inspected during its working life and regularly maintained. There was no sign of significant corrosion inside the twin box girders. In fact the seven layers of paint that protected the steelwork were in virtually pristine condition. And to blast off the paint in the areas to be inspected would mean parking compressors in the middle of the busy M5 motorway, with all the attendant disruption.
Time was also a major factor. 'We were supposed to have the bridge finished in time for the introduction of 40t lorries in 1999,' points out Highways Agency project manager John Bourne. 'A full survey would have taken up to six months - time we felt we couldn't afford.'
With the inevitable benefits of hindsight Bourne now says he wishes the Agency had taken the time and invested the money in a more detailed original survey, 'although it's not certain even that would have really warned us of the scale of the problem we are now facing'. Delays and costs have built up and the Agency has had to put in place new methods to make up lost time (NCE last week). Completition is now due by the end of 2000.
Strengthening the bridge was just one part of a contract to widen the 4km of M5 between Junctions 15 and 19. As it happened, the original designer, Freeman Fox, had built in enough width to the bridge deck to allow it to be widened from a symmetrical dual three lane layout to an asymmetrical dual four lane design (NCE 23 November 1995). Freeman Fox's direct descendant Hyder Special Structures drew up the plans for a complex strengthening operation and the Highways Agency went out to tender in late 1994.
'We were expecting tenders around the £80M mark so we were surprised when all the tenders submitted were close to £50M,' Bourne reports. 'In the circumstances we felt we were justified in accepting the lowest bid, which was from Costain.'
Around half of the £50M was earmarked for the strengthening works. Work started on site in June 1995. An immediate problem was a delay in access to certain sections of the road works due to overruns on the Second Severn Crossing approach roads to the north. Worse was to follow. As gritblasting operations began in the first 'northbound' box it soon became obvious that the thick paint had covered up a multitude of minor defects.
'We knew about all the major defects, mainly some fatigue cracking,' says Hyder Special Structures technical manager Philip Tindall. 'What we didn't expect was the sheer number of weld defects - stop/start faults, cratering, and oversize welds.'
Bourne adds: 'These didn't actually compromise the strength of the bridge, but we couldn't ignore them.'
Already faced with adding around 3,000t of extra steel in no less than 100,000 separate pieces, steelwork subcontractor Kvaerner Cleveland Bridge now had to cope with much more preparation work and customising of the new plates. Oversized welds where new plates were due to be added had to be ground down or the plates chamfered to fit around them. A more serious problem was found in a small minority of the 63mm thick plates used for the original diaphragms.
Tindall explains: 'When new plates were welded on, there were laminar failures within the original plate. This wasn't unknown in steel of the time.'
Further complications came after the completion of a Category III check on Hyder's design carried out by Frank Graham & Partners. 'This had more effect on the details of the smaller plates than the oversize welds,' says Hyder Consulting resident engineer Allan Chambers. 'Plates became smaller, the number of different shapes increased. Even though the total weight to be added is only 10% higher, the number of individual plates is up to 150,000.'
Some of the faults discovered, principally the delamination of the diaphragms, also needed redesign and Category III checks. All this added to the delays. and the cost.
However, the basic principles of Hyder's design remained unchanged. Temporary props relieve self-weight stresses over two adjacent access spans at a time. A fan of prestressed Macalloy bars between the top of the pier diaphragm and the webs reduces hogging moments and web shear stresses. Horizontal prestressed bars just below the top flange counterbalance added compressive forces in the bottom flange over the pier.
At the main river piers the boxes are haunched to a depth of 7.5m. Here, Hyder's design uses large tubular steel trusses laid on top of the bottom flange transfer stiffeners over the first third of the main and backspans. Cables run from each end of the struts over saddles near the top of the diaphragm, and are post-tensioned to induce a lifting moment in mid-span.
Three workfaces were planned, one in the river spans and one each in the approach spans. Bourne admits the steelwork subcontractor sometimes had difficulty finding enough welders and other skilled operatives to staff the workface teams. 'There is a national skill shortage at the moment,' he says, but insists the Agency would not have granted contract extensions for that reason. Be that as it may, the first really positive action to speed up operations was the introduction late last summer of three new squads inside the box girders.
These slot in between the gritblasters and the main workface teams, carrying out surveys and remedial work. 'In fact as time progresses the preparatory work gets further and further ahead of the main workfaces,' Tindall explains. 'Quite a lot can be done before the box is propped, such as cutting mouseholes and preparing new plates.'
The cross girders that connect the twin boxes can be strengthened independently of the props. Edge cantilevers also need strengthening to take new inspection gantries. Along the western edge of the northern span, moving the traffic out to the edge of the bridge requires new high containment parapets - which means even more strengthening to the cantilevers.
Although the Agency says it has been aware of the looming crisis for more than a year, it was October last year when the first signs of a major shake-up on site appeared. Following current Agency fashion, a partnering charter was signed on site and all parties sat down around the table to look at ways of speeding up progress. Chambers says the concerted application of value engineering techniques has already produced significant benefits.
'We now use studs instead of plug welds wherever possible, and we've increased the perimeter gap around the plates. This means increased weld size and other costs, but it's quicker and cheaper overall.'
The most dramatic sign of the new approach will be the introduction of an extra workface on the northern approach viaduct later this month. An extra set of the high-tech props has had to be ordered. This new work face is confidently expected to enable the site team to bring forward the revised completion date of March 2001 set last year to the end of 2000, Bourne says.
He also denies speculation that final cost will be as high as £150M or even £200M. 'We're holding discussions on a fair valuation for the extra work with Costain at the moment, but the final cost will be little more than £100M.' Of that, £70M will be for bridge strengthening, treb- ling the original sum, and the whole project will have taken nearly twice as long as planned.