Squeezed into a narrow and typically Pennine valley to the north of Todmorden is Lydgate Viaduct, an attractive structure comprising 13 spans. It forms part of the Copy Pit route (FHR6), a product of engineer John Hawkshaw, which traverses a geologically challenging landscape to reach Burnley, nine miles away. In doing so, the railway passes through Kitson Wood Tunnel – immediately east of the viaduct – and Holme Tunnel, further north, which Amco Rail partly reconstructed on Network Rail’s behalf in 2013-14. That intervention was prompted by a rotational landslip. We’ll recycle that sentence in a moment.
Attractive or not, it’s fair to say William Helliwell was not enamoured by Lydgate Viaduct. Immediately behind and rather overshadowed by it was Naylor Mill, which he owned. As the structure was being erected in the 1840s, Helliwell wrote to the Manchester Courier complaining that “various baulks, scaffolds etc etc so obstruct the light to my mill as to render it impossible for me to continue to run it with advantage either to myself or those in my employ.” Understandably exasperated, he closed the place down.
Months earlier, Helliwell had enjoyed temporary respite following the spectacular collapse of ‘Railway Mania’, an unsustainable investment frenzy which saw 272 Parliamentary Acts for new lines passed in 1846 alone. Many of those schemes were fraudulent, flawed or pointless duplications of rival routes.
In October 1847, activity on the ‘Burnley branch’ was halted – along with others being progressed by the Lancashire & Yorkshire Railway – bringing immediate redundancy for the many dozens of masons, carpenters and labourers toiling on the viaduct. Imagine the impact of that in an era before the welfare state. The winter had almost passed before the financial world had calmed sufficiently for work to resume, but with the economy measure that only one line of rails would initially be laid. It carried the inaugural train in November 1849.
A moving experience
We tend to forget that the environment around us is alive, even if it tends to move imperceptibly. The steeply-sided Calder valley has been contributing to the engineers’ workload for many years, resulting in the ongoing development of schemes to address a cracked retaining wall at Knott Road – just west of Lydgate Viaduct – and a slipping embankment in Kitson Wood, beyond the adjacent tunnel. Bridge reconstructions are also on the cards. But the viaduct itself exhibits the most eye-catching defects, not least because the main road runs right past it.
Movement of the structure was recorded between 1925 and 1934, as a consequence of which diagonal fractures developed in the western abutment’s curved wing wall. The parapet end on the north (Down) side was pushed inwards so the track alignment now includes a short transition length to ensure structure gauge clearance; a 20mph speed restriction has been imposed on the Down line. And then there’s the westernmost arch – Span 13 – which distorted significantly to its south elevation, becoming more Norman than Roman.
Over the years, remediation has taken place in the form of stitching and pinning, as well as the rebuilding of spandrel walls and parapets. There are mortar ‘tell-tales’ in the wing wall, dating back to the 1930s; several of these are cracked.
Investigations from 1949 concluded that the failures were caused by a rotational landslip, with the slip circle rising to the surface beneath a row of cottages. These had been pushed closer to the road and suffered heaving of their floors. It was recommended that material be removed from the north side of the abutment and the toe of the slope loaded to resist further movement. Neither recommendation saw action but the cottages appear to have been demolished between 1956 and 1961.
In the Seventies, tie bars and a steel frame were installed within Span 13, wedged around the arch barrel with bullhead rail laggings and supported by stepped concrete foundations at the foot of the pier and abutment. Twenty years ago, four more tie bars were inserted above Pier 12 to restrain a large bulge in the south-side spandrel wall.
Hide and seek
In 2005, a detailed examination of Lydgate Viaduct found it to be in generally fair condition, although no confined space survey was undertaken of its voided spandrels. This was progressed six years later in advance of the five-month blockade to facilitate the works at Holme Tunnel, together with investigations to confirm ballast depth and deck construction.
Using a CCTV camera, it was observed that – above Pier 12 – four voids contained loose debris (soil and stones) whilst a fifth, on the south side, was partly infilled with concrete. The Yorkstone roof slabs had fractured, settling into the small space below where they were resting on the fill material. On the abutment side of Span 13, the voided chambers had similar characteristics but most of the roof slabs remained intact.
The arch itself was found to be in poor condition. Although Aerocem pressure pointing had taken place in the past – giving the arch face a satisfactory appearance – very little mortar was present within the joints behind. Moreover, the bullhead rail wedges were severely corroded and the uppermost tie bar had broken. Very shallow ballast depth was recorded above the crown due to the distorted arch profile, resulting in the Up line sitting 100mm higher than the Down. It was clear therefore that the time had come for a more substantive intervention, progressed during the current Control Period.
Although some ground investigation data had been archived from previous studies of the hillside above the viaduct, Hyder Consulting (now Arcadis) commissioned BAM Ritchies and Datum Monitoring Services to conduct new investigations in 2014. These would inform the design and give advance warning of any ground movement.
The work involved the installation of an inclinometer within a 150mm diameter borehole sunk in rough ground on the north side of Span 13. The borehole log recorded medium/high cobble content, with soft-to-firm clay overlying sand to a depth of 5.5 metres. Below this, firm-to-stiff clay was encountered, becoming stiff at 16 metres and recovered as sandstone gravel/cobbles from 19.5 metres to 21 metres.
Ten sensors were provided to continuously monitor rotational movement within the wing wall, arch barrel and inner faces of the pier and abutment, whilst two further sensors check for crack propagation in the wing wall. All indications suggest that the viaduct is now stable, with only cyclical and seasonal movement.
J Murphy & Sons inherited the project as part of its Network Rail structures framework agreement, awarded at the start of CP5. The firm appointed Tony Gee & Partners as its design consultant.
The optioneering process was complicated somewhat by the viaduct’s Grade II listing. Amongst the approaches considered was the infilling of Span 13 with lightweight concrete – creating a voided, cellular structure – but this was ruled out because of loading implications. Instead, it was agreed that a sprayed concrete arch would be applied, springing off in-situ concrete walls and supported by capped piles.
Towards its eastern end, the viaduct straddles three roadways and a watercourse, with several buildings in close proximity. The western half crosses tree-covered rough ground that rises to the north-west. Running on the south side – just eight metres from the base of Pier 9 – is the main Todmorden-Burnley road. So space was at a premium and made all the more challenging by the six-metre change in levels between the highway and the 70m2 work area below Span 13, where all the key activity was focussed. Programming, therefore, had to mostly follow a linear progression.
Enabling activity began on site in November 2016, with the setting-up of a compound on a narrow strip of land across the road. On site, conditions were initially very wet, with water emerging from below the displaced section of wing wall. The situation was improved greatly by putting stone down and creating a new drain.
The immediate priority was to add weight to the toe of the embankment, preventing further movement; this involved clearing the vegetation, allowing excavations within which a gabion retaining wall was built. Thereafter, the existing slopes were benched, a berm installed and the ground regraded to form a route up to where a piling mat would be established. Access was also provided along the north side of the viaduct.
With wagons bringing in several hundred tonnes of material, regular contact was needed with the local authority to agree appropriate traffic management measures.
Hit and miss
The intention was to begin the main strengthening phase in January 2017, after the earthworks had consolidated, but a three-week delay was incurred due to local authority recovery works following flash floods.
Van Elle was sub-contracted to fulfil both the piling design and installation, which called for a 4×5 array of piles, each 18 metres deep and 450mm in diameter. On top of these, a nominal 650mm thick cap was to be cast, stepping up in three sections from south to north.
An auger machine was used at the outset, but consistently refused at a depth of 5-7 metres due to boulders being encountered, probably tipped there when a nearby cutting was first excavated. The methodology was then changed to an Odex rotary percussive system; however this was limited to 232mm diameter with a sacrificial case. To account for this, the pile depth was increased to 22 metres and the concrete fill around the inserted rebar cage changed to a grout.
An additional pile had to be sunk at the critical north-west corner due to the failure of No.1 pile which twisted and then refused at around 18 metres. Completion of the piling took about two months.
With the concrete poured for the pile cap, attention turned to the splay walls against the pier and abutment faces. It was originally anticipated that the necessary shuttering would be outsourced to a specialist contractor using an off-the-shelf system, but this proved impractical because of its cost and complexity. Instead, the design and assembly was delivered by Murphy’s in-house temporary works team.
Due to the shape of the splay and the quantity of rebar, neither a conventional concrete pour sequence nor the use of vibrating pokers was possible, instead driving the choice of a self-compacting concrete.
Spray that again
The spraycrete works to the arch – undertaken by Gunform – were generally progressed during the day, but with the initial critical elements programmed for overnight midweek possessions. These were also utilised by Ropetech Access Solutions for de-vegetation work, crack stitching and repointing in hard-to-reach parts of the structure.
Despite its poor condition, Network Rail was understandably reluctant to remove any of the existing steelwork in case this caused instability. However, its surfaces were cleaned and prepared, and general repairs undertaken to the surrounding masonry. It was then encased in a nominal 100mm regulating layer of sprayed concrete, holes drilled to allow the fixing of starter bars and a waterproofing membrane applied.
The main structural barrel was sprayed in layers of 75mm, allowing each one to gain a strength of 30N/mm² before progress was made with the next. The overall minimum depth is 475mm, with rebar added in two stages. On top of this – to improve the finish – is a 25mm flash coat.
On the south elevation, listed building consent required the application of stone cladding to soften the concrete’s visual impact. However, on the north side and inner faces, it has been left exposed to allow the structure’s evolution to be seen.
Earthworks improvement formed the concluding part of the project, dismantling the upper section of gabion wall after creating another berm in front of it. The whole embankment was then regraded – rising at 27 degrees from the road – and new toe drainage provided.
After nine months’ implementation, and many more in development, the all-in cost to Network Rail has been £1.2 million. That might seem a chunky sum for one concrete arch, but there’s so much more going on here that the casual observer simply can’t see. You can bet, though, that William Helliwell is looking down and quietly chuckling to himself.
Thanks to Mark Billington and Chris Atkins from Network Rail, and Murphy’s Dave Copson, for their help with this article.
This article was written by Graeme Bickerdike.