A road that took Grantham 20 years to build
Grantham's planners first sketched out the need for a southern relief road in December 2007, when the town was awarded Growth Point status — a government designation that came with ambitions for thousands of new homes and the infrastructure to support them. Nearly two decades on, that road is still not finished.
The project runs 3.5km, linking the A52 at Somerby Hill to the A1, and has been structured in three phases. Phase One, a new roundabout off the B1174, opened in August 2016. Phase Two, which extended the route to connect the B1174 with the A1 via a grade-separated junction, was completed in December 2022. Both phases were delivered, broadly on schedule, and both left Grantham's town-centre congestion essentially untouched — because neither connected to the town itself.
Phase Three is the section that changes that. It is the 'missing link' that completes the full route, and it is where Lincolnshire County Council, South Kesteven District Council, Homes England, and the Greater Lincolnshire LEP have found the project to be anything but routine. The costs have more than doubled from initial estimates; the completion date has slipped repeatedly; and the engineering at its heart — a 2,500-tonne bridge over both the River Witham and the East Coast Main Line — has had to be substantially redesigned.
Two separate crises have driven those delays. Understanding them means asking what it actually takes to push a 293-metre bridge across a live railway, millimetre by millimetre.
The engineering logic behind a bridge push
Crossing the East Coast Main Line is not like bridging an ordinary waterway. Network Rail operates one of Britain's busiest rail corridors — high-speed trains run at regular intervals throughout the day — and any construction directly above the tracks requires the railway to be shut down. Those shutdowns, known as possessions, are typically granted only at night, in windows of a few hours. Trying to assemble a 293-metre, 2,500-tonne structure section by section using cranes, in the dark, above live overhead electrification lines, is not practically feasible.
An incremental launch — the technique being used here, commonly called a 'bridge push' — solves this by doing the majority of the work away from the railway altogether. The bridge is assembled on the west bank of the River Witham on temporary supports, then advanced in stages called shunts using hydraulic jacks. The jacks move the structure forward a few millimetres at a time; the bridge slides into position rather than being lowered in from above. The railway remains operational for most of the programme, with the moment of crossing managed in tightly controlled sequences.
The technique demands precise control. Instruments measure several points on the bridge structure every second, and the team can halt the push immediately if any misalignment is detected. WSP designed the bridge; Galliford Try is the main contractor responsible for building it. The overhead electrification lines running beneath the intended path leave no margin at all for a trajectory error — a constraint that, as the project would discover, proved harder to guarantee than originally assumed.
When the ground gave way
Before the bridge itself became the problem, the ground beneath it was.
In July 2022, with Phase Three underway, site teams discovered a section of soft, unstable earth directly beneath the planned bridge alignment. Before any steel could be placed, the foundations had to be rebuilt from scratch — and at a depth that gave some indication of how serious the situation was.
Piling contractor PJ Edwards drove preliminary test piles first, to establish whether bedrock could bear the loads required. Once that was confirmed, the full programme began: 48 rotary bored piles, each 900mm in diameter, sunk to depths exceeding 36 metres below ground level across four piers. A further 109 piles of the same specification were then installed across two additional piers and one abutment. To put 36 metres in context: that is roughly the height of a twelve-storey building, driven downward.
The work was time-consuming and expensive, though the soft-ground costs are not separately quantified from the project's broader budget. What is clear is that the geotechnical intervention pushed Phase Three's programme back significantly — and set the stage for a second, larger disruption that would arrive two and a half years later.
The design error that added a year and up to £20m
The second crisis arrived in February 2025 — and unlike the ground conditions discovered in 2022, it could not be attributed to geology or unforeseen site circumstances. Engineers established that the original installation method was unsafe: under specific wind conditions, the structure risked moving out of lateral alignment during the push. No amount of careful monitoring would fix a method that was, at its root, incorrectly designed.
Lincolnshire County Council classified this as a supplier design error. The implications were immediate: work stopped, the push schedule collapsed, and the project faced at least a year's further delay. A redesign was required before any forward progress could resume.
The engineering response focused on increasing the number of intermediate supports beneath the bridge during the launch. According to published accounts, the pier count was raised from five to seven, giving the structure additional points of contact to resist lateral movement as it travels across the valley. More piers mean shorter unsupported spans at each stage of the push — reducing the degree to which wind loading can induce the kind of shift that made the original design unsafe.
The financial consequences were substantial. The project's base cost of £148m is now estimated to reach between £158m and £168m — a range that, taken against the original ambition of the entire scheme, represents roughly double what was first envisaged. Lincolnshire County Council has allocated £38.4m to support bridge completion and is pursuing legal and contractual action to recover the additional costs from the supplier. The current status of those proceedings has not been confirmed in detail publicly.
How you push 2,500 tonnes across a live railway
March 2026 marked the moment the project shifted from design table to live operation. With the redesign resolved and seven permanent piers in place, contractors began the bridge push — three planned shunts that will carry the 293-metre structure across the River Witham, over the East Coast Main Line, and beneath the overhead electrification lines that power trains along the route.
Each shunt is a discrete phase of the same incremental launch. The bridge advances until a pre-determined stopping point is reached; then the sequence pauses, temporary supports are adjusted, and the next shunt is prepared. At the time of writing — late June 2026 — the second shunt was underway, and the bridge deck had reached roughly one-third of the way across the valley. The three constraints it must negotiate arrive in sequence: the River Witham first, then the live railway span, then clearance beneath the catenary wires above.
That middle section — the East Coast Main Line — is where tolerances are tightest. Intercity services run throughout the operation, and the overhead electrification leaves minimal vertical clearance. The second-by-second monitoring the team established for this purpose exists precisely for this passage: any lateral or vertical drift that exceeds tolerance halts the push immediately, without discussion. Nothing about the sensing or the response protocol is precautionary theatre; it is the only way the operation can cross a live railway at all.
Once the deck reaches its permanent piers — expected in autumn 2026 — road-surfacing and approach works can begin in earnest, keeping the 2028 opening date within range.
What Grantham gets when the road opens
When the 3.5km route opens in 2028, Grantham gains a southern bypass carrying traffic directly between the A52 at Somerby Hill and the A1, bypassing the town centre entirely. That will matter to residents who have watched the project from a distance for most of its life.
The original purpose, set out when Grantham received Growth Point status in December 2007, was never simply to ease existing congestion. The road was designed to unlock development land and support housing growth — to make the town's expansion viable, not just bearable. The infrastructure was meant to precede and enable the demand it was built to serve.
Twenty-one years between designation and opening is, by any measure, a long time. Some of that reflects deliberate phasing across a complex scheme; a good deal more reflects what happens when ground conditions and design assumptions both turn out to be wrong in sequence. The original budget — now roughly half of what the project is expected to cost — suggests it was priced for the engineering problem that engineers initially believed they faced, not the one they eventually encountered. That is a different kind of observation from a generalised complaint about public works: both overruns here had specific, identifiable causes.
Whether the council recovers the costs attributable to the supplier's design error remains to be seen. The pursuit of that recovery does at least establish that responsibility for a specific failure has been assigned and contested through formal channels — which is how these mechanisms are supposed to work, even when outcomes are uncertain.