Excavation safety is moving in the wrong direction. In this latest article, Groundforce Shorco Engineering Director Oliver Smith examines why a 50% increase in excavation safety breaches cannot simply be attributed to poor weather. Instead, he explores the underlying issues driving the trend and outlines why stronger planning, engineering expertise and a renewed focus on safe excavation practices are essential to reversing it.
It’s convenient to blame the recent spike in excavation breaches on seasonal workload or bad weather. It’s also wrong.
A statistical analysis by the Building Safety Group last year flagged a 50 per cent rise in excavation breaches on construction sites (the finding was based on Health and Safety Executive data from 7,000 site inspections during the first half of 2025). What’s more, the group pointed to the fact that about 15 per cent of construction-related deaths involve excavation accidents.
So what’s behind this? Groundwater levels fluctuate. Rainfall happens every year. The industry has always dealt with those variables. The reality behind the rise in breaches is more uncomfortable: we’re not consistently controlling the conditions we design for.
“The industry doesn’t need more guidance – it needs more discipline”
Excavation support doesn’t fail in isolation. Most systems used on UK sites are robust, with significant safety factors designed into them – they will often show deformation before structural failure occurs. That should be reassuring. Instead, it has created a dangerous assumption that these systems are inherently safe, even when the conditions they were designed for are no longer present. That assumption is the problem.
In most cases, the failure does not sit in the design. It sits in the gap between what was assumed at design stage and what’s actually built on site. Ground investigations provide a snapshot in time and designers work to a defined brief. However, ground conditions are dynamic, sequences evolve and loading rarely remains within a neat set of parameters.
I’ve seen this first-hand. On one scheme, a propped excavation experienced a failure in a hydraulic waler system. While at face value it would have been easy to blame the equipment, the cause was a fundamental shift between design-stage assumptions and site reality.
Support frames were installed higher than specified and a critical dewatering system that formed part of the design basis had been switched off due to noise. Together, they fundamentally changed the conditions the system was designed for.
The failure didn’t occur because the system was weak – it was because the excavation no longer resembled the one that had been engineered.
Where responsibility lies
This is where procurement and delivery models come into focus. Temporary works for shoring are often provided by specialist suppliers, designing to the information they are given in a competitive environment.
If the brief assumes favourable groundwater, controlled surcharges or a fixed sequence, that’s what the design reflects. Expecting suppliers to automatically design for every possible scenario is unrealistic unless that envelope is clearly defined. The responsibility for that definition sits with the contractor.
But this is where pressure builds. Tight programmes and commercial constraints mean temporary-works co-ordination is too often treated as a process to get through rather than a control to rely on. The temporary works co-ordinator, in theory, holds a position of authority. In practice, that authority is often diluted the moment it conflicts with the programme. The result is predictable: excavation continues even when conditions have already changed.
The industry doesn’t need more guidance. It needs more discipline. This starts with treating the design envelope as a live constraint, not a fixed document. Before excavation begins, there must be a clear point where a competent person verifies that site conditions still match the design intent. If they don’t, the work stops.
This isn’t about a managed slowdown – it’s about a hard stop to realign site reality with engineering intent. Anything less than a full halt is simply managed risk, and that’s where the real danger lies.
There’s growing interest in digital monitoring: sensors, live data and automated alerts. These tools have their place, particularly on complex schemes. But they’re not a substitute for understanding the ground. Technology can tell you that something has changed. It can’t replace the judgement required to understand why.
If there’s one bad habit that has become normal on UK sites, it’s this: continuing to excavate when something has already changed. If that approach persists, breaches will continue, regardless of the season, weather or equipment used.