The same project, two very different verdicts
In April 2026, the Grantham Southern Relief Road project received a Gold Award at the Considerate Constructors Scheme National Site Awards — the UK's highest independent recognition for construction-site conduct. Contractor Galliford Try and the project team were judged among the very best sites in the country for safety culture, environmental management, workforce wellbeing, and community engagement.
For anyone who has watched this road fail to open for years, that award may feel like a strange headline. This is, after all, the same project that halted a 2,500-tonne bridge launch in February 2025 when designers discovered a critical wind-loading error — and the same project that had to redesign its central bridge from scratch back in July 2022 after unstable ground was found beneath the eastern embankment. The combined cost of those two failures runs to an estimated £10–20m above an already-revised budget, with the road now not expected to open until around 2028.
Both facts are real. Neither cancels the other. Understanding how that is possible reveals something important about how large infrastructure projects are actually built — and governed.
What the CCS Gold Award actually measures
Established in 1997, the Considerate Constructors Scheme scores construction sites against independently assessed criteria: health and safety, environmental performance, workforce wellbeing, site appearance, and engagement with local residents and stakeholders. Achieving Gold — the scheme's highest tier — means an independent monitor judged the GSRR site among the very best in the country on those dimensions.
The achievements behind that verdict are specific and genuine. Over one million tonnes of waste materials were recycled on site. Six jobs were created for previously unemployed local people, thirteen work-placement students and five apprentices came through the project, and more than 200 students were engaged at career events. Schools received £6,000 in donated time and materials. Local charities received more than £7,500. These are not tokenistic add-ons; they are precisely what the scheme exists to recognise and reward.
What the CCS does not assess is equally worth stating plainly: delivery performance, budget adherence, and design quality fall entirely outside its scope. A project can be two years late, tens of millions over budget, and twice interrupted by engineering errors on its central structure — and still win the highest national site-conduct award. That is not a loophole. Site conduct and engineering design are governed, assessed, and procured through separate institutional channels, and the CCS makes no claim to judge the latter. The award is legitimate; the system is simply narrower than most people assume.
First failure: the ground that wasn't accounted for
Summer 2022 was when the project's first serious trouble surfaced. Site teams working on the phase-three bridge — the structure that would carry the new road over both the East Coast Main Line and the River Witham — found that the eastern embankment was sitting on soft, unstable ground. The original foundation design had not accounted for it.
Lincolnshire County Council was direct in its public statement about where responsibility lay. The co-ordination of ground-investigation data and the subsequent design of the bridge foundations was, the council said, 'the responsibility of the bridge designer, WSP.' The implication was plain: WSP had the information and the design remit, and the condition had not been adequately reflected in the foundations it specified.
The fix required more than patching. Because the ground beneath the eastern side was too soft to build on, the entire bridge had to be redesigned to route around the problem. Two additional piers were added, extending the structure from roughly five spans to approximately seven. In practical terms, that meant longer steel sections, revised loadings, a new planning application, and a significantly extended fabrication timeline. A bridge that had already been designed had to be substantially redesigned.
The original target completion date of late 2023 was effectively lost at this point. The exact cost of the 2022 failure in isolation was not confirmed publicly — the council acknowledged at the time that the changes would add 'several million' to the budget, but no precise figure was published.
Second failure: the wind nobody modelled
Three years after the ground problem forced a redesign, the project reached what should have been its most dramatic moment of progress. In February 2025, crews prepared to push-launch the completed bridge — 293 metres long, 16 metres wide, weighing approximately 2,500 tonnes — across the live East Coast Main Line and River Witham below. The structure, manufactured by Briton Fabricators, was to be nudged incrementally into position on temporary supports over one of the UK's busiest rail corridors.
The attempt was halted before it concluded. Designers had failed to model how crosswinds acting on such a long, slender structure during the launch process could cause it to flex and drift out of alignment — a safety-critical omission on a live railway where millimetre precision is not optional. Lincolnshire County Council stated publicly that work stopped because they were 'told it couldn't be pushed into place as intended due to concerns relating to specific wind conditions.' Safety, the council said, came first.
What compounded the embarrassment was the process that had already been applied. According to the council, all design work on the project had passed through a mandatory independent checking layer — the safeguard specifically intended to catch errors of this kind before they reach site. The checker did not catch it. The identity of the checking firm has not been made public.
Rectifying the failure is expected to cost between £10 million and £20 million, bringing the project total to approximately £158–168 million — roughly double the original budget. Legal and contractual recovery processes were initiated. Push-launch work eventually restarted in March 2026, some 13 months after the aborted attempt.
The principal-agent gap large projects cannot close
Councillor Richard Davies put the structural problem plainly in the council's 2025 statement: 'The complexities of designing and constructing this relief road, particularly this bridge, are far beyond what we as a council can do directly.' That sentence is the clearest available expression of a challenge that runs through large infrastructure delivery across the UK — not just in Lincolnshire.
The logic that follows from it is straightforward. When a client lacks the specialist knowledge to evaluate the work, it procures that knowledge from outside. Then, because it still cannot verify the expert's output independently, it procures a second layer of experts to check the first. On the GSRR, that is precisely what happened: WSP was engaged as bridge designer, and all design work passed through a mandatory independent checking process before reaching site.
The problem is what occurs when both layers fail — not simultaneously, but sequentially, on entirely separate failure modes. The 2022 ground failure was geotechnical: a question of whether the foundation design properly reflected available ground-investigation data. The 2025 wind-loading failure was structural-dynamic: a question of whether the chosen launch method modelled how a long, slender structure behaves in crosswind during incremental pushing. These are different engineering disciplines and different analytical blind spots. The checking layer missed both.
This is the principal–agent gap in its most direct form. The client commissions expertise it cannot evaluate, then commissions oversight of that expertise — and if the checker shares an assumption with the designer, or simply fails to interrogate what the designer did not model, the client has no internal recourse. Davies's observation that 'mistakes still occurred' despite the checking layer captures the limitation with inadvertent precision.
The CCS Gold Award sits entirely outside this chain of accountability — which, as established earlier, is precisely why the two verdicts can coexist without contradiction.
What Grantham's road teaches about building anything hard
The bridge is moving now — a few millimetres per minute over the live East Coast Main Line and River Witham, its 2,500 tonnes held steady while crews maintain the alignment precision that the February 2025 attempt could not. Completion of the 3.5km A1-to-A52 link is projected around 2028. The original rationale for building it — to relieve congestion through Grantham — has not changed, and the road will do that work when it opens.
What the project leaves behind, beyond the road itself, is a clear and reusable illustration of how complex infrastructure actually fails. The GSRR did not fail because Lincolnshire made poor choices or because WSP is uniquely negligent. It failed because two entirely distinct engineering problems — one geotechnical, one structural-dynamic — each defeated the governance layer designed to catch it. That is a different lesson from 'things went wrong.' It is a lesson about the limits of delegation and the limits of checking.
For anyone watching a large infrastructure project — here or anywhere — the practical takeaway is not cynicism but precision. A well-run site and a technically troubled project are not contradictions; they are assessed by different institutions against different criteria. And when a project has an independent checking layer, asking what the checker was specifically tasked to verify is not a hostile question. It is the right one.