At a glance
The short version
- An EM depth is computed from how field strength changes between antennas, assuming one straight, isolated conductor with a clean signal. 1Jump to source 12Jump to source 2
- A GPR depth is computed from travel time and an assumed wave velocity — and velocity changes with the material. 3Jump to source 3
- A sonde depth is the depth of the capsule, read from a dipole field, not from the pipe's crown. 4Jump to source 4
- Manufacturer accuracy figures describe ideal conditions. Our receiver's spec says ±3% depth precision — under suitable conditions — and the qualifier is the load-bearing phrase. 1Jump to source 1
- Survey-grade GNSS sharpens where the mark is. It does nothing for how deep the line is. 6Jump to source 6
- Report depth as an estimate with its basis, never as a clearance. Depth changes; the excavator's care must not.
_No locating instrument measures depth. Every one of them calculates it from something else — field geometry, travel time, null spacing — and every calculation rests on assumptions the ground is free to violate._
Where each number comes from
Electromagnetic. The receiver compares signal strength across vertically separated antennas; the ratio implies distance to the conductor. The mathematics assumes a single, straight, horizontal conductor, isolated from other fields, carrying an undistorted signal. Under those conditions the manufacturer specifies ±3% depth precision for our receiver, with locate accuracy of ±5% of depth. 1Jump to source 1 Every congested-corridor condition from the distortion guide — coupled parallel lines, bent fields, bleed-off — violates an assumption and moves the number, sometimes far beyond the spec, with no warning printed on the screen. Ground composition, frequency and signal strength all shift real-world readings. 2Jump to source 2
Ground-penetrating radar. The radar measures two-way travel time. To turn time into depth it must assume a wave velocity, and velocity is a property of the material: it differs between dry sand and wet clay, between native soil and trench backfill, between old concrete and new. 3Jump to source 3 Calibrate against a target of known depth and the number tightens; carry one velocity across a site of varying moisture and it quietly drifts. A GPR depth is only ever as good as the velocity assumption behind it.
Sondes. A sonde's depth comes from its dipole field — read directly in sonde mode, or approximated by measuring the spacing between the two nulls and multiplying by 0.7. 4Jump to source 4 Two honest offsets ride along: the reading is to the capsule, which rides the pipe invert, so a large-diameter pipe's crown is shallower than the number; and a tilted sonde (in a sag or on a slope) skews the field pattern the calculation assumes.
The error budget, itemized
Think of every depth reading as carrying a stack of debits:
| Assumption | How it breaks | Typical effect |
|---|---|---|
| One isolated conductor | Parallel utilities share the signal | Depth reads shallow or deep, unstable between passes 2Jump to source 2 |
| Straight horizontal run | Tees, bends, risers, dips near the reading point | Geometry error right where readings get taken — at fittings |
| Undistorted field | Re-radiation from neighbors, rebar, fences | Peak offset and depth error together 1Jump to source 12Jump to source 2 |
| Known signal | Weak current, high frequency bleeding off | Noise inflates the calculation 2Jump to source 2 |
| Known velocity (GPR) | Moisture and material change across the site | Systematic drift, worst in mixed fill 3Jump to source 3 |
| Level capsule (sonde) | Tilt in sags and slopes | Skewed nulls, wrong 0.7 spacing 4Jump to source 4 |
| Reading equals cover | Grade changed since installation | The line didn't move; the surface did |
That last row deserves a sentence. Depth is measured from today's surface. Regrading, paving lifts, erosion and landscaping all change cover without touching the pipe — a reading taken this year describes this year's grade.
What good practice looks like
- Take depth where the geometry is clean — mid-run, away from fittings, risers and crossings — never directly over the meter because it happens to be convenient.
- Read twice, from opposite directions. Agreement earns confidence; disagreement is a distortion flag, and the flag matters more than either number.
- Sanity-check across methods when it matters. An EM depth and a GPR depth arrived at independently and landing together are worth more than either alone; apart, they tell you an assumption is broken somewhere. 3Jump to source 3
- Calibrate GPR velocity when a depth actually matters — against a known target, a measured cleanout, or an exposure on the same site. 3Jump to source 3
- Record the basis with the number. "42 inches, EM peak mode, mid-run, stable both directions" is a professional depth. "42 inches" is a liability wearing a tape measure's clothes.
What depth is for — and what it is never for
A depth estimate is planning information: which crossing needs potholing, whether a bore has any business at that elevation, how much caution the first two feet deserve. It is not a clearance. Under ASCE/UESI/CI 38-22, depth from surface geophysics belongs to Quality Level B designation; a depth you can design against comes from Quality Level A — the utility exposed, surveyed, and seen. 5Jump to source 5 Where a project's vertical clearance actually governs, the honest sequence is: designate, pothole at the critical point, and survey what the pothole reveals.
And one distinction our own deliverables lean on: the RTK GNSS riding on our instruments records where the mark is to centimeter-class positions per the manufacturer's specification — it contributes nothing to how deep the line is. 6Jump to source 6 A map can carry survey-grade horizontal positions and estimate-grade depths at the same time, and an honest legend says exactly that.
The excavator's rule survives every paragraph above: treat every depth as approximate, expose by hand or vacuum within the tolerance zone, and let the pipe — not the display — have the last word.
Practice drills
- Read depth over a known, exposed line — then re-read five feet away, at a fitting, and near a parallel run. Log the spread; that spread is your personal error budget.
- On one GPR line, process the same hyperbola with two plausible velocities and watch the "same" target change depth. Then calibrate against a known target and see which velocity the site actually supports. 3Jump to source 3
- Depth a sonde by receiver reading and by the 0.7 null-spacing method on the same setup; reconcile the two before trusting either in the field. 4Jump to source 4
View 6 numbered sources
Sources
Every numbered claim above traces to one of these. Links go to the publisher of record.
- RD8200 locator specification — Radiodetection — https://www.radiodetection.com/sites/default/files/product-downloads/rd8200-ts-eng.pdf Source for the ±3% depth precision and ±5% locate accuracy figures and their conditions qualifier.
- Utility locator depth accuracy: what changes the reading — GIS User — https://gisuser.com/2026/07/radiodetection-utility-locator-depth-accuracy-test-what-changes-the-reading/ Source for ground composition, frequency and signal-strength effects on real-world depth readings.
- ASTM D6432-19 — Standard Guide for Using the Surface Ground Penetrating Radar Method for Subsurface Investigation — ASTM International — https://www.astm.org/Standards/D6432.htm Source for velocity-dependent depth calculation and material variability.
- Locating a sonde — Radiodetection Support — https://support.radiodetection.com/hc/en-gb/articles/360001077883-Locating-a-sonde Source for sonde depth reading and the 0.7 null-spacing method.
- ASCE/UESI/CI 38-22 — Standard Guideline for Investigating and Documenting Existing Utilities — American Society of Civil Engineers — https://ascelibrary.org/doi/book/10.1061/9780784415870
- Reach RS2+ technical specifications — Emlid — https://emlid.com/reachrs2plus/ Source for RTK horizontal accuracy as a position property, distinct from target depth.
Before you use any of this
This is educational writing, not training. Reading an article does not make anyone competent to locate, and nothing here qualifies, certifies or authorizes you to perform this work. Competence in this trade comes from supervised field experience under someone already qualified — on real sites, with real equipment, making real mistakes where somebody senior is watching. There is no substitute for that, and this page is not one.
The risk is real. This work happens around energized electrical conductors, pressurized gas, and open excavations. Getting it wrong injures and kills people. A mismarked line, a misread depth, a ghost signal taken for a real one — each of those has put someone in a hospital or a grave. Treat every line as live and every interpretation as unverified until it is proven otherwise.
Follow the authorities that actually govern your work, not this article: your employer's procedures, the equipment manufacturer's instructions, the facility owner's requirements, applicable OSHA and state regulations, and the direction of the qualified or competent person on site. Where any of those differ from anything written here, they govern and this does not. If you are not qualified and authorized for a task, do not attempt it — and if a situation is outside what you have been trained for, stop and escalate rather than improvise.
Sourcing. This is original writing by High Desert Scan & Locate, LLC. It cites third-party statutes, standards, codes and publications, all of which remain the property of their respective owners. We describe them in our own words and do not reproduce their text. Where a standard governs your work, obtain the current edition from its publisher and read it — the published source controls; our paraphrase does not.
No representation or warranty. Provided "as is." We make no representation or warranty, express or implied, as to accuracy, completeness, reliability or currency, and this is accurate only as of the review date shown, if at all. Nothing here is legal, engineering, surveying or safety advice, reading it creates no professional or client relationship, and nothing here clears any area for excavation, cutting or coring.
Company names, product names and trademarks are used for identification only and do not indicate affiliation, sponsorship or endorsement. We receive no compensation for any mention.
Corrections: info@highdesertscan.com. See our Editorial policy, sourcing and attribution.
Private utility locating supplements, and does not replace, public one-call (811) notification. Utility designation is Quality Level B per ASCE/UESI/CI 38-22 unless otherwise stated. GPR interprets subsurface conditions; it does not expose them. Results are interpretations of conditions present within the requested area on the service date, and depend on target material, depth and site conditions.