Service strikes are still one of the most common and costly incidents on UK sites. A permit-to-dig is not a formality; it is the control that stitches together detection, supervision, method and change management. When it’s done well, the team knows exactly where not to dig, where to hand-expose, who is watching the bucket, and when to stop. When it’s rushed, the first scrape becomes a gamble with power, gas, fibre, and programme.
TL;DR
/>
– Plan the dig from drawings and detection, not from memory or guesswork, and brief the team at the trench.
– Mark services clearly, set hand-dig zones, and use trial holes or vacuum excavation where risk is high.
– Keep the permit live at the point of work; stop and re-issue if conditions, people, or scope change.
– Put a competent spotter on the excavation, slow the plant, and treat the last metre as hand-dig only.
– Record what you find and update the service plan before you close the permit.
Permit-to-dig controls that actually work
# Stage 1: Define the dig and gather information
/> Start by pinning down the exact footprint, depth and purpose of the excavation. Pull utility plans from the undertakers early and overlay them with site drawings, temporary works and recent service diversions. Assume records are incomplete and that services may be deeper, shallower or offset. Walk the route and note street furniture, joint boxes, meter cabinets, valve covers, and recent reinstatements that hint at live runs. Identify interfaces: deliveries, pedestrian routes, plant movements, and adjacent works that could distract or crowd the dig. Build the risk picture before anyone lifts a spade.
# Stage 2: Detect and mark services
/> Use a calibrated CAT and signal generator to sweep the route in multiple modes, with a competent operator and a witness. Where congestion, ground conditions or critical services demand it, bring in GPR or vacuum excavation as added layers. Mark services on the ground with clear colour codes, arrows for direction, and written depths if confirmed. Photograph the mark-up and capture it on a simple sketch so the knowledge survives rain, mud and foot traffic. Set out no-go zones and hand-dig areas with barriers or paint boxes, not just a verbal note.
# Stage 3: Build the method and issue the permit
/> Write a method that matches the detection outcome, not a generic template. Define plant type and bucket, approach direction, lift plans if lifting covers or slabs, and how spoil will be stored without loading edges or burying markings. Specify hand-dig tools (non-conductive where appropriate), trial holes at key crossings, and vacuum excavation where strike energy must be eliminated. Name the supervisor, plant operator, spotter and service watch; brief them at the trench with the permit in hand. Time-limit the permit, link it to a daily start-up check, and state the stop-work triggers. PPE sits last in the hierarchy, but ensure dielectric gloves, eye protection and arc-rated clothing are considered where the risk justifies it.
# Stage 4: Break ground under control
/> Verify the permit and mark-up at the point of work before starting the machine. Hand-expose to prove positions, then keep the machine back until lines are positively identified. Use a dedicated spotter who is watching the bucket and the markings, not running errands; slow the slew and keep bites shallow near known routes. Maintain a clean excavation edge with physical stop-lines; keep spoil away from markings and edges. Maintain exclusion around the plant, control banksman/plant communications, and stop when visibility drops in bad light or heavy rain. Treat any uncertainty as a pause for re-detection, not as a nudge forward.
# Stage 5: Manage change and close out
/> If an uncharted service appears, or a known one doesn’t, stop. Make the area safe, notify the supervisor and asset owner, and update the permit. Re-brief if the crew, plant, or route changes. Record what you actually found: depths, offsets, caps, ducts and conditions. Only close the permit when the excavation is backfilled or made safe, photos are filed, and drawings or service registers are updated for the next phase.
A short, real-world scenario
/> A groundworks gang on a housing spine road needed a 600 mm deep trench to pick up a new street lighting duct. Plans showed LV running on the opposite verge, so the supervisor kept the machine in and aimed to hand-dig crossings only. The CAT sweep was done early morning, but by the time the 13-tonner started, rain had washed most markings and a delivery wagon was nudging cones to get past. On the third bucket, the operator felt a nudge and stopped; a black cable appeared at 350 mm, offset from the drawing. The permit had technically expired at midday, but no one noticed. They isolated the area, brought in vacuum excavation to expose the full length, and found an old diversion that never made it onto the latest plan. The close call led to a tighter rule: re-validate the permit at lunch, re-mark after rain, and hand-expose every crossing before the machine goes anywhere near it.
Supervisor’s pre-dig checklist
/>
– Confirm utility plans are on site, understood, and matched to today’s location and chainage.
– Verify CAT/Genny is calibrated, the operator is competent, and the sweep is documented and witnessed.
– Walk the line with the crew; re-mark services and set hand-dig zones and physical barriers.
– Check plant type, bucket, exclusion zones, and that a dedicated spotter is named and briefed.
– Agree stop-work triggers: uncharted services, lost markings, change of crew/plant, poor visibility, water ingress.
– Line up the right tools: non-conductive spades, insulated matting if required, vacuum kit where specified.
– Keep the permit at the point of work; sign on crew and re-validate after breaks or weather events.
Common mistakes
# Treating old utility plans as gospel
/> Record drawings are a starting point, not a guarantee. Services migrate over time, get diverted, or sit at variable depths.
# Permits written in the office, not at the trench
/> If the method and signatures aren’t agreed where the work happens, you’ve missed the point. Briefings must be face-to-face at the dig.
# Rushing the last metre with the machine
/> Most strikes happen close to the target. Make the final approach a hand-dig or vacuum task, even if the programme is tight.
# No one owns the cable avoidance kit
/> If everyone assumes someone else has checked the CAT, nobody has. Nominate a person, record checks, and keep the kit with the team, not in a distant store.
Keeping discipline once the trench opens
# Cable-strike prevention priorities for the next 7 shifts
/>
– Pin up today’s permit and mark-up at the trench and refresh markings after rain or vehicle scuffing.
– Rotate a single competent spotter role per shift so attention doesn’t drift; relieve them on breaks.
– Stage vacuum excavation for known crossings and congested zones instead of “seeing how it goes”.
– Photograph every trial hole with a tape on depth and offset; file to the job folder before close of play.
– Escalate any uncharted find immediately to the site manager and service coordinator; pause downstream works until the route is proven.
Bottom line
/> A permit-to-dig works when it turns unknowns into knowns, puts a named supervisor on the trench, and slows the job at the point of highest risk. Keep the paperwork live, the markings visible, and the bucket honest, and you’ll avoid the strike that wrecks both safety and programme.
FAQ
# When do we actually need a permit to dig?
/> Issue one whenever you break ground, including small trial holes and fence posts near suspected services. It’s good practice on refurb, fit-out externals, civils, and even soft landscaping where buried utilities may be shallow. The permit links detection, method and supervision into a single control, which is why it should not be skipped for “quick” tasks.
# How long should a permit remain valid?
/> Tie validity to the shift and conditions rather than a blanket date. Re-validate at the start of each shift and after significant changes such as heavy rain, loss of markings, crew changes or altered scope. If you move location or depth, close it and issue a new one with a fresh briefing.
# What’s the minimum we should do for service detection?
/> As a baseline, use a calibrated CAT and signal generator with a competent operator, and record a multi-mode sweep. Where drawings are poor, routes are congested, or the consequence is high, add GPR or vacuum excavation to remove uncertainty. Always prove positions with hand-dug or vacuum trial holes before bringing in the machine.
# Who should operate the CAT and Genny?
/> Someone trained, practised, and named on the permit, ideally accompanied by the supervisor to witness the sweep. Avoid leaving it to whoever is free; inconsistent technique gives inconsistent results. Store the kit with the team and keep a simple check log for calibration and battery status.
# What should we do if we uncover or strike an uncharted service?
/> Stop, make the area safe, and keep people clear; do not attempt to move or repair it. Notify the site manager and the relevant asset owner and follow the escalation route in your RAMS and permit. Treat it as a change: update the mark-up, re-brief the team, and only restart when the revised method is agreed and the permit is re-issued.
