Cable strikes remain one of the most stubborn, high-consequence risks on UK projects. Whether you’re trenching for drainage, piling near a footpath, or installing bollards on a refurbished frontage, you can’t rely on “should be clear” thinking. The expectation is simple: prove where services are before you dig, then control the dig so nothing drifts. That means evidence, not assumptions, and a permit that ties the proof to the task on the day.
TL;DR
/>
– Get all utility info early, but treat drawings as clues, not proof. Prove on the ground with detection and trial holes.
– Use the right tools for the mix of services: locator and signal generator, and GPR where needed. Mark, measure and photograph finds.
– Tie it to a live permit-to-dig. Define hold points, stand-off distances, and who can authorise changes.
– Keep proving as you go: brief the team, supervise plant, hand-dig to confirm, and stop if anything doesn’t match expectation.
Prove-before-you-dig: a staged controls playbook
# Define the excavation and the service risk
/> Pin down exactly where and how you’re going to break ground. Mark the extents, depth, and plant to be used. Note interfaces such as footpaths, kerbs, existing chambers, lighting columns, comms cabinets and boundary walls that often betray buried assets. This is the baseline for what services could sensibly be present.
# Collect and reconcile information
/> Pull utility records, client as-builts, diversion drawings and any previous survey results. Don’t assume they align; flag mismatches and date everything. Aim for a clear statement of confidence: “we have credible info for X routes, gaps for Y, and suspect unknowns in Z.” If records are old, incomplete, or the area has been remodelled, escalate for enhanced detection and programme the time to do it properly.
# Detect using the right methods
/> Decide the detection approach to suit the risk. A cable locator and signal generator can pick up metallic services; ground-penetrating radar can identify non-metallic utilities and buried structures. Use multiple passes, different frequencies, and both passive and active modes. Map and colour-mark suspected routes, and record anomalies for targeted trial holes rather than guessing.
# Validate on the ground
/> A mark on tarmac is not proof. Validate with trial holes in the right places, using hand-digging or vacuum excavation to expose and identify. Confirm depth, orientation, and type, then tag and protect what you’ve found. If nothing is found where the survey suggests, treat that as a serious finding: pause, re-scan, and widen the search before releasing plant.
# Mark and control the dig zone
/> Transfer the verified locations to durable on-ground marks that will survive weather and traffic. Take dimensioned sketches and photos showing measurements to fixed features so the information remains usable if paint fades. Set physical barriers, signage, and no-go zones to keep plant out of protected corridors. Make it obvious to a new starter where they can and cannot cut, core or trench.
# Permit-to-dig and brief the team
/> Issue a task-specific permit that stitches all of this together for the day of the dig. Include the verified routes, the stand-off distances, hand-dig-only zones, trial hole positions, and hold points for supervisor sign-off. Brief the crew and the machine operator together, with the sketch in hand. Confirm insulated tools are available and used where needed, and that everyone knows who can vary the plan (hint: not the excavator on their own).
# Dig safely and keep proving
/> Proceed in short sections and keep a competent person on the ground watching for changes. Hand-dig to confirm service positions ahead of plant and maintain sensible clearances that match your briefing. If you encounter uncharted ducts, concrete, or backfill different to expectation, stop and reassess; don’t “just edge around it.” Keep plant and attachments controlled, use a banksman in tight areas, and manage spoil so it doesn’t bury your marks.
# Record, sign off, and update service plans
/> As you expose services, capture their actual positions. Photos with tapes to kerbs or columns, simple redlines to drawings, and notes on depth are enough if done consistently. When the section is complete, close out the permit, keep the evidence with the site records, and update the service plan for the next team. Good records shorten the next dig and avoid repeating the same uncertainty.
Scenario: weekend push on a housing site service trench
/> A civils gang is installing a new water connection along a cul-de-sac within a housing development. Friday’s programme review shows a utility survey from months ago and a note about “probable LV” along the footpath, but the team is under pressure to open the trench on Saturday to meet a road closure window. The supervisor walks the route and sees a new EV charger pedestal near Plot 12 that isn’t on the drawing. He arranges a re-scan late Friday: the locator picks up a strong signal drifting into their planned trench line, and GPR shows a shallow duct crossing. The permit is held, and the first metre on Saturday is hand-dug to confirm; they expose a twin-duct with comms and LV 200 mm shallower than expected. The line is remarketed, stand-off increased, and the dig sequence altered so the trench is stepped and the live services remain protected. The trench still opens on time, but without the near-miss that would have made Monday’s meeting very different.
Supervisor’s quick-check list
/>
– Are all available utility records and surveys on hand, current, and consistent with site conditions?
– Has the right detection been used for the service mix, and have suspect gaps been escalated?
– Are service routes clearly marked on the ground, with dimensioned sketches and photos taken?
– Have trial holes confirmed positions where the dig comes close to expected routes?
– Is a live permit-to-dig in place with hold points, hand-dig zones and permissions defined?
– Is supervision in place to control plant, maintain clearances, and stop work if conditions change?
Common mistakes that keep causing strikes
# Treating a drawing as proof
/> Drawings show intent or history, not today’s reality. Always prove on the ground before plant turns a track.
# Assuming GPR sees everything first time
/> GPR interpretation is skill-dependent and conditions vary. Use it alongside locators and validate with targeted trial holes.
# Rushing the trial hole
/> A quick scrape with a shovel is not verification. Expose fully enough to see orientation, depth, and whether multiple services are bundled.
# Letting plant drift into no-go zones
/> Banksman distracted, line fades, bucket creeps. Keep physical barriers or timber baulks, refresh marks, and reset the briefing if the plan moves.
Short-term site focus
# Next 7 days: make the invisible visible
/> Pick one live dig and test how solid the “proof” really is. Walk the line with the service sketch, the permit, and a locator; confirm that marks still match what the instrument reads, and that photos and measurements are usable. If anything is unclear, pause and bring in detection support or hand-dig to verify before plant proceeds. Build the habit of closing the loop with redlines and photos at the end of each shift so the next crew inherits certainty, not guesswork.
Bottom line on proving services
/> Proving services is not a paperwork exercise; it’s a sequence of detection, validation, control and supervision that stands up when the bucket bites. When time gets tight, the only workable shortcut is better evidence, not wishful thinking.
FAQ
# When is a permit to dig needed and what should it include?
/> Use a permit whenever there’s a possibility of underground services or when control over the dig needs to be formalised. It should link verified service information to the exact task: marked routes, hold points, hand-dig-only areas, stand-off distances, and who can approve any change. Include the briefing record and contact details for escalation. Expiry should be short so you’re not relying on stale info.
# What counts as proof that a service is located?
/> Proof means you can show where it is with confidence, not that you hope it isn’t there. Typically that’s a combination of detection records, on-ground marks, photos with measurements to fixed features, and trial holes confirming position. Keep it proportionate to risk: the closer the dig, the stronger the proof needed. If you can’t show it, you haven’t proved it.
# How close can plant excavate to a known service?
/> There isn’t a one-size figure that suits all sites. Set a sensible clearance in the method and permit based on voltage, service type, ground conditions, and plant control, and enforce hand-digging where appropriate. Use physical means to maintain separation, such as timber buffers or barriers. If conditions change, reassess the clearance before continuing.
# What if detection results are unclear or conflicting?
/> Treat uncertainty as a stop signal, not a footnote. Re-scan using different modes, bring in additional methods like GPR, or expand the search area. Targeted trial holes can resolve conflicts quickly if done carefully. Record decisions and, if necessary, re-sequence works rather than pushing on blind.
# Who should carry out service detection and verification?
/> Detection should be done by competent people using calibrated equipment and with time to interpret results properly. On smaller jobs, that might be a trained supervisor and an experienced operative; on higher-risk digs, bring in specialist surveyors. Verification by trial holes should be led by someone who understands what they’re looking for and when to stop. Always brief the wider crew so the knowledge isn’t stuck with one person.
