Unrecorded duct and a cracking trench
You are excavating a 1.4 m deep trench in the verge of a 30 mph residential road to connect a new street lighting column. Utility plans show an 11 kV cable crossing the trench line at about 0.8 m and a 300 mm water main running parallel 1.5 m away; no telecoms are shown. You have swept the line with a CAT and Genny and marked the signals in paint. At 0.9 m the operative exposes a duct tile that appears on no plan, at a point where the CAT gave no signal. The trench has been open since the previous afternoon. The sides are vertical in firm clay, which has begun to crack and spall near the top, and there is 200 mm of water at the base. Spoil is heaped about 500 mm from the trench edge and the mini-excavator has been tracked to within 1.5 m of the edge to reach the next bay.
1. Set out what should happen in the next 30 minutes in relation to the duct.
Read the worked answer
Stop mechanical excavation along the whole run and treat the duct as live and unidentified until proved otherwise.
The correct inference from 'no plan record and no CAT signal' is not 'nothing there'. It is that the plan is incomplete and the instrument did not couple to that particular service - small ducts, plastic water pipe, some fibre and services at depth routinely give weak or absent responses. The detection method has failed once on this trench, so its null results carry no weight for the rest of it.
In practice:
- No machine digging within the hand-dig zone around the duct and around any paint-marked service; hand-dig only, with insulated-handled tools, and a safe digging technique (shallow scraping, no vertical striking).
- Re-scan with the Genny in its different modes (power, radio, inductive) and with a second detector, and consider a trial hole or ground-penetrating radar before going further.
- Identify the owner from the duct type and location and contact them before disturbing or, if damaged, before backfilling.
- Brief the whole gang on the change; the person who found it must not be the only one who knows.
- Record the find, its depth, offset and the actions taken, and update the drawing used for the rest of the job.
- If the duct is damaged or exposed in a way that could be a danger, isolate or protect it and treat the area as a live-services zone until the owner confirms.
2. The clay is cracking and the trench has stood overnight. What does the support and inspection regime need to look like?
Read the worked answer
A vertical-sided trench at 1.4 m in drying, cracking clay that has stood overnight is not a stable excavation - it is an excavation that has not yet failed.
Reasoning: the cracks and spalling near the top are tension features in the desiccated crust; they reduce the shear strength of exactly the zone that provides the resistance to the sides falling in. Water at the base softens the toe. Time is the fourth factor - the trench has been open while nothing was inspected or recorded.
So:
- Do not send anyone back in until a competent person has assessed it and suitable support, or a safe batter, is in place. For an excavation of this depth in these conditions, the default expectation is a proprietary trench box or sheeting and struts, not an assumption that clay stands on its own.
- Formal inspection regime: by a competent person at the start of each shift, after any event likely to have affected stability (rain, vehicle loading, a fall of material, an adjacent excavation), and recorded. The absence of a record since yesterday is itself the finding - treat the trench as unassessed.
- Control the water - pump or drain it, and stop it running back along the trench line.
- Decide whether the trench should be open at all overnight. Unattended open trenches need covers or secure barriers and should not be left if they can be backfilled or boxed.
3. What is wrong with the spoil and the machine position, and how does that feed into the decision about the trench?
Read the worked answer
Both are loading the trench from behind the failure plane.
Spoil at roughly half a metre from the edge is, in practical terms, sitting on the crest of the excavation. It adds vertical load exactly where the clay is already cracked, it removes the option of battering back, and it provides loose material that will fall in if the side gives. The guidance position is to keep spoil and materials back from the edge by a distance related to the depth - for a trench of this depth, a metre or more clear is the working assumption, and more if the ground is poor.
The mini-excavator within 1.5 m is worse: a tracked machine transmits dynamic load and vibration into the ground beside the trench, and vibration is one of the recognised triggers for a spoil-side collapse in desiccated clay. It also encroaches on the only route out of the trench if something goes wrong.
Practical response: stop tracking alongside the trench; move the machine to a defined standing area and, if the work requires it close, use a properly designed load-spreading measure or a trench box that is rated for the surcharge. Relocate spoil to a designated heap. Then re-inspect before anyone re-enters. The spoil and machine are not housekeeping issues - they are part of the stability calculation and belong in the assessment of whether that trench is safe at all.