The two Tredington-Ashchurch and Grove Lane M5 overbridges were built in 1969/70. Both bridges comprise four span continuous in situ reinforced concrete voided slabs, supported on skeleton abutments with encastre pier columns. Deck articulation is by pier flexure with elastomeric abutment bearings. Foundations are spread footings in Lower Lias Clay.
Checks undertaken as part of the Highways Agency's bridge assessment and strengthening programme showed the unusually long and slender piers required strengthening. Reinforced concrete jackets were designed by Halcrow to improve column strength.
Following a competitive tender process, a contract for the work on both bridges was awarded to Norwest Holst Construction under a conventional ICE 5 contract. The 14-week contract began in February last year.
Three days after excavation began at the Tredington-Ashchurch Bridge, buried surfaces of the columns showed softening that could easily be removed by hand.
This damage became more severe with depth and concern grew over the strength of the columns as the extent of softening, apparently attributable to sulphate attack, reached the reinforcement in places. A continuous core taken through the remaining fill to the base showed this was also affected.
As a precaution, works were temporarily suspended and the minor road over the bridge was closed to traffic. Detailed testing of the columns was carried out. At first this simply included taking incremental dust samples to check chloride and sulphate levels, with core samples taken for petrographic examination and compressive strength testing by materials consultancy STATS.
A sample of clay backfill was also tested for sulphate content and pH. Although these initial tests showed a high level of sulfates (32.5% as SO3 by mass of cement in the first 30mm of concrete), the minimum core strength was 61.0 N/mm2. Limestone coarse and fine concrete aggregates were predominant in the concrete.
Soil tests indicated Class 2 sulphate conditions in accordance with the Building Research Establishment Digest 363. With such good quality concrete in a soil of Class 2, it was soon apparent that this was not the form of sulphate attack normally encountered.
The possibility that it was thaumasite sulphate attack (TSA) was suggested by Halcrow head of material technology David Slater, who was aware of the work carried out by BRE. The discovery was soon confirmed by BRE using X-ray diffraction
While these investigations were being carried out, the Agency and Halcrow were considering the structural implications. Column bracing was required to allow full depth excavation and reopening of the road over the bridge, albeit with a 3t weight limit.
With the columns braced the excavation was completed to reveal the full extent of the sulphates. This showed the lower parts of the columns, together with the tops and sides of the base, to be severely damaged by TSA.
As the extent of sulphate attack became evident investigations were widened to include mapping, surface profiling, drilling and rebound hammer testing. Further core samples were obtained for petrographic analysis, X-ray diffraction and scanning electron micro- scopy, to establish the type and cause of the sulphate attack. Cores were taken from the columns and bases.
Soil samples were taken at various distances from the faces of the columns, by window sampling undertaken by Soil Mechanics before large-scale excavation.
Testing also revealed high chloride levels at the depths of the reinforcement in the TSA-affected parts of the columns. A few smaller areas were discovered where significant corrosion of the reinforcement had occurred.
Cores taken through the base showed the underside to be unaffected by TSA, although limited thaumasite formation was founded in pre-existing voids within 10mm of the interface with the clay. Top surface of the foundations was affected by TSA to a maximum depth of 35mm.
At the time of the TSA discovery BRE advised that even concrete beyond that affected could still deteriorate, due to the
presence of absorbed sulphates and the tendency for crystalline growth of thaumasite in the presence of moisture.
This, and the need to expedite the works that were affecting motorway traffic flows, led to a decision to temporarily support the deck and completely replace the lower parts of the TSA-affected columns. Working within the shored excavations, the damaged concrete on top of the base was removed by water jetting to what was judged to be a conservative depth beyond the limit of TSA formation.
The concrete was re-cast and an impervious membrane used to separate existing construction from new concrete without limestone aggregate.
The changes caused by the TSA discovery enabled the bridge articulation to be altered. This facilitated reduction of work to the central reserve pier and minimisation of traffic disruption. The work undertaken at Grove Lane Bridge was similar to that for Tredington-Ashchurch.
Norwest Holst was given details of the new construction on 1 April 1998 and asked to mobilise resources. RMD was appointed to design and supply temporary supports and proposed Megashor towers with a capacity of 1000kN per leg. Halcrow revised the support conditions in the LEAP5 assessment model to calculate jacking loads and check load effects in the deck.
In April vertical loads were transferred to the temporary supports at Grove Lane Bridge, with load transfer at Tredington-Ashchurch bridge three days later. Work on the verge piers was completed in July, just before the busy summer period for M5 traffic when the Agency required all lanes to be open.
Work on the central jackets restarted on 7 September with Lanes 2 and 3 closed in both directions. These had a programmed completion date of 23 December but the revised centre pier design and contractor's acceleration brought this forward to 20 October 1998.