In the Upper Clyde Valley the river snakes between hills and the West Coast main line (WCML) crosses it four times. The four span 101-metre Lamington viaduct is the largest of these crossings. It is all very picturesque, and the sound of the river burbling around the bridge piers is usually quite restful.
However, when Rail Engineer visited on 22 January, the noise from the River Clyde was much more pronounced and the ambience anything but restful. After overnight rain and melting snow, the Abington river gauge station, six miles upstream from Lamington, was recording the river level as 1.8 metres above datum, compared with its average 0.7 metres. Flooding of the site compound car park showed the rising water level.
Storms Desmond, Eva and Frank resulted in the wettest December since records began and created what the Centre for Ecology and Hydrology (CEH) described as “extraordinary” hydrological conditions. On the day of storm Desmond, river water discharge into the seas around Britain was a third more than the previous maximum. Published CEH statistics show that, during the month, the Clyde’s river flow was a record 249% of its long-term average.
On 30 December, storm Frank brought widespread flooding and disruption to Scotland. The M74 motorway was blocked by a flood at Abington, where the gauge station recorded the Clyde at its highest-ever level – 3.12 metres above datum.
Tale of three decks
The original viaduct was part of the Caledonian main line between Carlisle and Glasgow which opened in 1848. It was a bow spring truss structure, built on three masonry piers. In 1937 and 1938, the viaduct was re-decked with a plate girder structure for each line. This kept both lines open during construction while the Up line deck was built on downstream concrete pier extensions. It also gave a larger radius curve enabling the Coronation Scot service, introduced in 1938, to run at higher speed. With this new arrangement, the upstream end of the original masonry piers no longer supported a bridge deck.
Sixty years later, there was a requirement for a further increase in speed. This time it was to 125 mph for the West Coast upgrade. So in 1999, the old deck was removed and replaced with one having reinforced main beams and a concrete deck. This was slid into position onto the existing piers during a long weekend disruptive possession.
Despite the problems elsewhere, on 30 December there was no disruption to trains between Carlisle and Glasgow. However, at 07:35 the following morning, a train reported a dip in the track over the viaduct. In response, a temporary speed restriction (TSR) was imposed pending the arrival of track maintenance staff who, after examining the track and watching a train pass over the viaduct at low speed, lifted the TSR. They remained on site and, after a few trains had crossed the bridge safely at speed, they observed unusual track movement as a later train passed – a few minutes later the line was blocked.
The Rail Accident Investigation Branch is to investigate this incident as a dangerous occurrence (as defined by the Reporting of Injuries, Diseases and Dangerous Occurences Regulations). In addition to considering the response to the notification of a track dip on the viaduct, the report will examine the effectiveness of Network Rail’s processes to mitigate risks to structures in extreme weather.
An emerging problem
The extent of the underwater damage to the bridge was not immediately apparent on the 31 December. After river levels had fallen further, a photograph of pier two taken on 1 January shows a single sandstone block missing. That this was at the top of a pyramid of missing blocks would only become apparent from a diver’s inspection.
On New Year’s Day, monitoring points were also set up which showed pier two had settled by 125 mm. There was not settlement at the other piers. Rock armour was placed at pier two to reduce flow and limit scour damage. The following day, a diver from Subsea ROV & Diving Services inspected the piers. His report showed the pier two scour damage to be far worse than expected. Below the waterline most of the masonry pier’s sandstone blocks had been washed away. Although the concrete pier was intact, there was void below it. The footprint of pier two’s foundation had been reduced by three-quarters and offset from its centre line.
With this asymmetric support and the heavy load on the foundations (1,500 tonnes for the deck and 1,000 tonnes for the pier’s weight), it became clear that there was a possibility that further scour might result in the collapse of the viaduct.
An initial date of 1 February was given for reopening the line, but this was revised to the first week in March when the full extent of the damage was known following extensive investigation works.
Securing the viaduct
AMCO (Amalgamated Construction) has both the framework and emergency response contracts for structures in Scotland with Network Rail. Infrastructure Projects route delivery director Stewart MacPherson advised that AMCO was mobilised on New Year’s Day and they also got a local quarry opened for the rock armour. By 3 January, work had started on an emergency haul road to deliver material for the construction of a causeway upstream of the viaduct.
With the risk of collapse, an exclusion zone was declared to prevent anyone working on or below the viaduct. As a further precaution, overhead lines were cut before and after the viaduct as its collapse might have brought these wires down onto anyone working close to the viaduct. Not surprisingly, full advantage of this unplanned closure has been taken to work on the line.
This risk was mitigated, to an extent, by the rock armour upstream of the pier and construction of the causeway. Yet the viaduct could only be considered secure when mass concrete had filled the void below the piers. Before this could be done, causeways had to be built on either side of the viaduct to dam the river between the piers and get materials to the working area. These included reinforcing bars and one-tonne sandbags needed for rough shuttering. Due to the exclusion zone, these sandbags were positioned by 360° tracked excavators at long reach to ensure their cabs were not under the viaduct. Constructing the causeways required 1,500 tonnes of stone.
On 8 January, all was ready for the first mass concrete pour. This used a boom pump which did not require anyone to enter the exclusion zone under the viaduct. Sixty cubic metres was poured into the void that day or, as Stewart put it, the equivalent of a portacabin. In the event, filling the void took four and a half days and three hundred cubic metres of concrete (five portacabins) to fill the void. By then, the pier had settled a total of 155mm, 30mm since the line had been blocked. Once the concrete had set the viaduct was secure and the exclusion zone was lifted.
Pier two took the brunt of storm Frank’s record water flow, in part because of a bend in the river just before the viaduct. Its permanent repair requires a reinforced concrete jacket on either side of the pier as designed by Donaldson Associates. Before the full extent of the scour damage was known, it had been thought that a jacket would only be required on the damaged south side of the pier. Each jacket has a footprint of 21 x 2 metres and rests on the concrete poured to secure the viaduct. The jackets are being fixed into position by eight metre soil anchors of which 75 are required on the south jacket and 50 on the north.
Pier two also requires new bridge bearings as the settlement has caused them to exceed their travel limit. To restore the viaduct to its original level, the bridge deck is to be lifted by 160mm using the concrete jackets as jacking points. It is hoped that new bearings will be fitted during this operation.
However, Stewart explained that it may not be possible to procure the three bearings before repairs to the viaduct are complete. This will not delay re-opening as a temporary bearing solution is being developed for this eventuality. If required, the new bearings could be fitted during a weekend rules of route possession.
Piers one and three
Pier one was undamaged but is to have additional rock armour which will be angled in line with the river flow. This is to be done at all three piers.
There was no settlement at pier three. However, there was some scour damage which required fifty cubic metres of concrete to repair. On 10 January, the masonry at the nose of pier three collapsed. It had experienced a significant increase in flow due to rising river levels (1.3 metres at the Abington gauge station) and the causeway was directing the entire river around it. With no deck on its upstream end, this collapse did not affect the viaduct’s structural integrity. However, it will require a new nose end to protect the remaining part of the pier.
It had been recognised and accepted that construction of the causeway would increase the water flow around pier three as the priority was to protect pier two and get concrete poured into its base. Stewart considered this to be, literally, a race against time. Following its collapse, the west river bank was re-aligned to reduce the flow past pier three.
At the time of writing it was estimated that the viaduct will not re-open until the first week in March, closing the line between Glasgow/Edinburgh and Carlisle for two months. Train operating companies are supportive of Network Rail’s decision to close this route until it is safe to run trains. For example, ScotRail has reduced its local services to Kilmarnock to allow Virgin Trains to run an hourly shuttle service between Glasgow and Carlisle via Dumfries, with a stop at Kilmarnock for ScotRail passengers.
Clearly there is a requirement to open the viaduct as soon as possible. Passengers want to know when they can travel without disruption and Network Rail’s compensation payments to TOCs for this unplanned disruption are almost certainly greater than the cost of the viaduct repairs.
The difficulty for Stewart’s team was that it took some time to determine the full extent of the work required. After the end-of-January completion date was announced, it was found that the concrete required to stabilise pier two was five times that expected, pier two required two concrete jackets and replacement bearings and the nose of pier three collapsed. The completion date was thus revised to the first week in March. This was announced at a press conference on 17 January and was widely reported in the Scottish media.
Although the scope of the work is now fixed, the programme is still affected by changes in river level. At the time of Rail Engineer’s visit on 22 January, the causeway was under water and the reinforcing bar frame for the concrete jacket footing was just visible above the water level. No concrete was poured on that day. To mitigate the effect on the programme of varying river levels, an alternative access to pier three was prepared on the north river bank, although this was not required.
There was further flooding on 26 January when the Abington river gauge recorded the Clyde to be 2.26 metres above datum. This flood washed away the northern riverbank and damaged the viaduct’s north abutment, requiring further emergency repairs.
Work continued in difficult working conditions from storms Gertrude and Henry. By 4 February, the bridge deck at pier two had been raised to its original height. This required eight jacks which were supported by the new concrete jackets. The central bearing plinth was then demolished to break out the old bearings. This work was completed by 6 February.
The replacement custom-made steel bearings were delivered earlier than expected. They were installed over the weekend of 13/14 February after a concrete plinth to support them had been constructed. Following final repairs, including work on the track and overhead line, the viaduct re-opened to traffic on Monday, 22 February, two weeks ahead of the previously announced March date.
A question of resilience
The trains that are now crossing Lamington viaduct are doing so over a structure designed to withstand extreme weather events such as storm Frank. Yet the closure of a key rail route for seven weeks inevitably raises questions such as the one asked in the Scottish Parliament on whether there was a winter resilience plan for Lamington viaduct. In response Scottish Transport Minister Derek Mackay advised that the viaduct was covered by Network Rail’s resilience planning and that no concerns had been highlighted with this structure.
The plan Mackay referred to is the route Weather Resilience and Climate Change Adaption (WRCCA) plan produced by Network Rail Scotland. This includes rail asset vulnerability and impact assessments and includes a programme to calculate the risk of bridge scour damage.
Lamington viaduct’s last annual detailed inspection was in March. This included an examination for scour by divers which revealed no issues. Up to then, the viaduct’s masonry piers had withstood the River Clyde’s flow for nearly 170 years. Yet a few weeks ago they were unable to do so. Reasons for this are likely to include the cumulative effect of December’s extreme storms, the way the river flows around the piers immediately after a river bend before the viaduct, as well as other factors yet to be assessed.
Both the RAIB report and Network Rail’s own review will doubtless identify lessons to be learnt from Lamington so that Network Rail’s WRCCA plan can be revised accordingly. However, with the unprecedented weather conditions, it remains to be seen whether the severe scour damage was reasonably foreseeable. What is predictable is that extreme weather events will continue to test the resilience of Britain’s rail infrastructure.