Top 7 ways underground
services get hit
The damage is obvious. The cause isn’t always.
A call and paint on the ground are not enough. The current before-you-dig process is falling short. Count down seven leading causes in the Common Ground Alliance’s (CGA) 2024 US/Canadian damage reports, then look at the pressures behind them.
#7of 7The locate request went unanswered.
The asset owner or its locator received a valid request. The crew started digging before a response came back: no locating, marks or reply.
The waiting is real. Across eight US 811 centres studied by CGA, 38% of standard requests still had incomplete responses when work was due to start. CGA warns that unpredictable timing can erode confidence in the process. CGA report, pp. 21–22.
#6of 7Paint on the ground. No physical check.
The crew dug before potholing to check the marks, or failed to keep checking as the excavation progressed.
Paint is a clue, not an X-ray. Potholing takes time and money. Australian interviews describe time and cost pressure pushing locating and checks down the list. That’s where “get it done” can become “take the risk”. Australian research, pp. 22–23.
A line copied from a plan is still a plan. At Augview, we’ve seen services marked “APP” (“as per plan”) where the position has not been confirmed on site. That is plan information, not a verified locate.
#5of 7“Other excavation error.” That narrows it down…
This groups digging mistakes that don’t fit the report’s more specific categories.
“Other” is a roomy filing cabinet. It ranks fifth as a group, not as one specific mistake.
#4of 7A confident mark. An incorrect location.
The service was marked in the wrong place. The reported cause was locator error.
A crisp paint line can still be confidently wrong. Australian interviews point to differences in locator skills, equipment limits and tight deadlines. Confidence is not confirmation.
The records themselves can be uncertain too: only around 60% of surveyed BYDA users agreed they could rely on the supplied information’s accuracy. Australian research, pp. 12–13 and 24–25.
#3of 7The service was there. The mark was missing.
A service was missed during locating and marking. The work area may have been misunderstood, or the request marked as clear by mistake.
If the plan shows a service but the ground has no marks, find out why. “No marks” needs an explanation before the next street is asking why there’s “no Netflix”.
#2of 7Found it. Checked it. Hit it anyway.
The service had been exposed and its position checked. Then the excavation equipment got too close. The crew failed to keep enough clearance around the service.
Finding it is one job. Keeping the equipment clear through every pass is another. A checked position is not a finished job. As the excavation moves, the operator still needs a clear picture of where the service runs.
#1of 7No locate request was made.
The biggest category starts with a skipped step: nobody told the local before-you-dig service they planned to dig. That’s 811 in the US, with regional services in Canada. It appears more often than inaccurate marking and missed-mark locator errors combined.
In NZ and Australia, the enquiry is free, but on-site locating can cost. beforeUdig and Before You Dig Australia pass the enquiry to participating asset owners. Some NZ owners also charge for plans. The crew still needs locating and physical checks to suit the job.
Why skip the process? Time, cost and gaps in awareness. Australian research identifies cost and schedule pressures that can push locating down the priority list, alongside poor awareness among homeowners and smaller contractors. Good records and skilled locators only help when the job brings them into the picture. Australian research, pp. 22–25.
The seven, side by side.
Missing or misplaced marks (#3 and #4), verification (#6) and keeping clear (#2) account for about 41.5% of reports with a known cause in CGA’s 2024 US/Canadian data. These are areas where crews need clear information and a shared understanding of the site.
The plan has to make sense on site.
The plans can reach an inbox and still miss the crew. Australian interviews describe plans not reaching subcontractors at handover. Australian research, pp. 22–23.
A locate that never updates the records leaves the next crew with the same problem. Australian interviews describe records staying unchanged even when someone offers the findings to asset owners. At Augview, we also see returned data that does not fit the owner’s database format. The findings need to return in a form the owner can use. Australian research, pp. 20–21.
Remove the surface and you remove the paint. Marks can also fade or be lost as work progresses. CGA records lost or unmaintained marks separately from the seven above: 2.17% of reports with a known cause. A useful site picture needs to survive longer than the paint. CGA report, p. 17.
At Augview, our goal is to help crews connect the records with the ground in front of them. We put mapped services into the camera view of the site, so crews can compare records with field findings, update asset records as they work and talk through the same picture with clients and colleagues.
With suitable records aligned to the site, Augview gives crews a shared view to use alongside locating, physical checks and agreed excavation controls. See the layout. Discuss what differs. Update what you find.
Show the crew the mapped layout on site.
See how Augview puts mapped services into the site view, helping crews and clients understand the layout. Talk through where it could help on your jobs.
Request a free demoSee how Augview worksSources, definitions and how we ranked the seven
We selected the seven largest counts from CGA's published 2024 top-ten table and count backwards to #1. The report analyses 25 known-cause categories; seven is the article's selection, not the full list. Its sample covers 196,977 unique reported damages in the US and Canada. Percentages are the publisher's known-cause shares, which exclude unknown/other entries; they are not percentages of all 196,977 reports.
The categories describe reported predominant causes, not complete investigations. Numbers 3, 4 and 5 are broad categories. The local sources reviewed do not provide an equivalent seven-cause ranking. This does not mean local records do not exist; their coverage and categories differ. No geographic correlation is assumed.
The Australian explanations draw on BYDA’s commissioned research, including 135 service users, 78 asset owners and 20 stakeholder interviews. They describe reported experiences and contributing pressures, not a local ranking or explanations proved for every CGA incident. APP and database-format observations reflect Augview’s industry experience; their prevalence is not measured here. The 38% measures incomplete responses across eight US 811 centres; the 41.5% groups four damage categories, not measured results from Augview. Australian research · CGA timing study.
The request growth is BYDA’s reported 2024/25 year-on-year change. The CGA trend is a modelled US damage index using consistent reporters, not a strike rate per enquiry: 100 in 2022, 94 in 2023 and 96.7 in 2024. The 2.9% rise is calculated as (96.7 ÷ 94 − 1) × 100. These series are not combined. The PHMSA figure comes from the DOT-hosted snapshot retrieved on 2 October 2026, filtered to 507 unique gas-distribution incidents with cause “EXCAVATION DAMAGE” and incident years 2010–2024; INJURE and FATAL fields sum to 133 and 23. It is not a current all-utilities total or evidence of injuries rising every year. DOT snapshot catalogue · PHMSA reporting definitions.
For the local referral process and fees: NZ service FAQs · Australian service FAQs.
Read the source and method notes · Download the displayed data
Original records and calculation notes
The seven-cause ranking
CGA’s 2024 findings on EIN Presswire (webpage) list the ten leading causes and their percentages. The full CGA report (PDF, hosted by Copperhead Industries) supplies the event-count table on printed p17. We preserve its first seven, in reverse order for the countdown. The unit is a unique reported damage event, not cost or likelihood per excavation. CGA’s original report host displayed a security check during our link test; the links here provide the CGA-authored announcement and the inspected report copy. The sample contains 196,977 unique damages (189,052 US; 7,925 Canada), from voluntary confidential reports.
The seven counts sum to 111,838. That is 56.8% of all 196,977 damages. The publisher's seven rounded known-cause percentages sum to about 77.5%; unknown/other cases are excluded from that percentage denominator. The exact known-cause total and complete 25-category appendix were not accessible, so we do not invent an unknown-only segment or recalculate the published percentages.
Categories 3, 4 and 5 are broad. The DIRT reporting guide (PDF) was used to check the definitions. Readable headings paraphrase the labels; the download retains their exact source categories.