A permit-to-dig system only earns its keep when it stops spades and buckets from meeting live services. On UK sites, strikes still happen in plain sight: rushed enabling works, vague drawings, late design changes, “it’ll be fine” hand-digging, and multiple gangs working the same corridor. The result is rarely just a broken cable. It’s electric shock risk, gas release, flooding, service outages, programme chaos, angry neighbours, and a site leadership team trying to explain why the basics weren’t in place.
This is a practical guide to making a permit-to-dig process work on the ground—so excavations, trial holes and breaking out are controlled, supervised and communicated in a way that suits real site pressures.
What a permit-to-dig is actually controlling (not just paperwork)
Permit-to-dig is there to control uncertainty. Drawings, records and utility plans are useful, but they’re not “truth”. Services get diverted, abandoned assets remain in the ground, and depths change across a few metres. The permit is the point where you force the job to slow down long enough to confirm what’s there, agree the method, and set boundaries on the work.
A solid permit-to-dig system should lock in:
– Where the work is allowed (a defined area, not “somewhere over there”).
– What methods are allowed (machine dig, vacuum excavation, hand dig, breaking out).
– How buried services are located and protected (scan results, trial holes, spotting).
– Who is in charge (a named issuer and a named supervisor accepting it).
– When it expires (shift-based control, not an open-ended permission slip).
– What changes kill the permit (weather, plant swap, route change, new subcontractor, redesign).
This is not about making every small trench a drama. It’s about stopping uncontrolled digging from becoming the default.
How it plays out on site: a short scenario that feels familiar
A civils gang is cutting a new duct trench across a trading estate access road on a tight weekend closure. The foreman has utility plans printed from an email chain, and the banksman has a CAT and Genny in the back of the van. By late afternoon, the surfacing crew is asking for a bigger excavation so they can hit level and get a clean reinstatement before Monday morning traffic. A different machine turns up because the planned 3-tonner has gone to another job, and the new driver hasn’t worked this site before. The supervisor is pulled away dealing with delivery wagons and complaints from a neighbouring unit about blocked access. The gang “opens it up” with the bucket to save time, and the first sign of trouble is a smell and a hiss from the bottom corner of the cut. The whole closure is suddenly about emergency calls, evacuation, and who actually authorised the change.
That chain of events is exactly what permit-to-dig should break—by making scope, plant, supervision and methods non-negotiable unless the permit is re-issued.
One workflow that works: from information to controlled excavation
Start by treating “dig” as any intrusive ground work: excavation, trial holes, coring, piling near services, breaking out slabs, even driving pins or stakes. Consistency prevents loopholes.
# Step 1: Pin down the location and interfaces
/> A permit should reference a sketch or marked-up plan with a clear boundary. On complex sites, split into zones so permits don’t become so big they’re meaningless. If the dig sits near live traffic, site welfare routes, or existing temporary works, make those interfaces visible on the permit and on the ground with spray marks, barriers, and signage.
# Step 2: Use detection properly (and treat misses as information)
/> CAT scanning is useful but not magic. It’s operator-dependent, ground-condition dependent, and easy to compromise when people rush. Good practice is to scan the whole area, not just the line you hope to dig, and to record what was found in a way that the next shift can understand.
When nothing shows up, don’t translate that into “clear to dig anywhere”. Translate it into: “We still need trial holes and controlled excavation until proven otherwise.”
# Step 3: Trial holes and positive identification
/> Trial holes are where you turn uncertainty into facts. They should be planned, supervised and done with the correct method (often hand-digging or vacuum excavation) until services are confirmed and their runs understood. Treat trial holes as a task with its own controls: edge protection, spoil placement, access/egress, and a plan for water management so holes don’t become hidden hazards.
# Step 4: Set the digging method and the “no-go” rules
/>
This is where permits often go vague. Be specific about:
– Machine digging limits (depth/offset, approach direction, and when to stop).
– Hand dig zones around identified services.
– Use of insulated tools where appropriate.
– Required spotter/banksman presence.
– Whether breakers are allowed (and under what conditions).
– Physical protection once services are open (supports, bridging, sandbags, plates—whatever suits the utility and the exposure).
If the method won’t fit the programme, the answer isn’t “just be careful”—it’s to re-plan the work, bring the right kit (like vacuum excavation), or change sequencing.
# Step 5: Supervision, handover, and permit closure
/> Permits should be shift-based and tied to a named supervisor physically present. If the supervisor changes, the permit should be re-accepted, not waved through on a phone call. Closing out should confirm: services left protected, backfill completed to spec, markers/tapes installed where needed, and the area left safe for other trades and the public.
Permit-to-dig checklist for the supervisor’s pocket
Use this as a quick pre-start filter before anyone breaks ground:
– Confirm the dig boundary is marked on the ground and matches the permit sketch/plan.
– Ensure scan results are recorded and briefed, including “no detection” areas and assumptions.
– Identify how the first 300–500mm will be opened (hand/vac) and who is doing it.
– Verify plant, driver and spotter match what the permit allows (any change triggers stop/re-issue).
– Set exclusion zones and traffic/pedestrian controls, including spoil placement and access routes.
– Agree the stop points: service found, unexpected material, water ingress, smell/hissing, unstable edges.
Common mistakes that keep causing strikes
# Treating the plan as proof
/> Utility drawings and as-builts are a starting point, not confirmation; overconfidence leads to machine excavation where a cautious approach is needed.
# Permits issued for “the area” rather than the task
/> A wide permit becomes a blanket permission, and people drift beyond the safe boundary as soon as the job “needs a bit more room”.
# Last-minute plant swaps without re-briefing
/> Different bucket sizes, machine weights and operator habits change the risk instantly; if the permit doesn’t reflect the plant, it’s not controlling the work.
# Scanning done once, then ignored
/> Marks fade, the workface moves, and conditions change; without re-marking and ongoing supervision, the scan becomes a tick-box stay-behind.
Keeping momentum without shortcuts: practical controls that don’t slow the job
A good permit-to-dig system supports production by removing surprises. A few site-realistic habits make the difference:
Use physical marking that survives a shift. Spray paint alone disappears under traffic and mud; back it up with stakes, pins, barriers, or offset marks that remain visible. Where the public interface is tight, set a clear corridor with cones and signage so you’re not constantly moving barriers to let people through.
Make the trial holes the programme driver. If the plan is “we’ll dig and see,” you’ve already accepted chaos. Plan trial holes early, get services positively identified, then release machine digging where it’s genuinely safe.
Hold a two-minute permit pause when anything changes. Weather turning ground to slurry, a new driver on site, a surfacing crew pushing for a wider excavation—these are common change triggers. The stop doesn’t need a meeting room; it needs authority and a pen.
Protect exposed services like you expect someone to forget they’re there. Coverings, supports and barriers stop the “quick step across” becoming a trip, crush, or snag incident once other trades arrive.
A one-week service-strike prevention push
# Seven days to tighter excavation control
/>
1) Map every active excavation and upcoming break-out on a single site plan, then pin it on the board for all supervisors to work from.
2) Re-brief plant operators and gangs on the non-negotiables: first dig method, stop points, and who can authorise changes on the day.
3) Re-mark existing scan lines and boundaries with more durable indicators where traffic, rain or spoil is wiping out paint.
4) Introduce a simple “permit expires at handover” rule so shift changes and supervisor swaps trigger re-acceptance, not assumptions.
5) Walk the site with civils, M&E and temporary works leads together to spot pinch points where multiple trades will push each other into unsafe digging.
Permit-to-dig isn’t about perfect paperwork; it’s about making uncertainty visible and controlled before the ground is opened. In the next briefing, ask three questions: Who is actively supervising the first dig? What change would stop the job immediately? And what will still be in place—markings, protections, barriers—after the gang has moved on?
