How to run and document a fall-of-potential ground test
Most ground test reports that get rejected aren't rejected because the number was high. They get rejected because nobody can tell whether the number is real. A single reading with no probe distances, no plateau check and no soil conditions proves nothing — the inspector or engineer of record can't tell whether the probes were 30 feet out in a parking lot full of rebar or 150 feet out in open ground. This template makes the method part of the record: probe distances, the 52/62/72% readings, the plateau check, soil and weather, instrument calibration, and a sign-off.
The fall-of-potential method (IEEE 81) in plain terms
The three-point or fall-of-potential method is the standard acceptance test for a new grounding electrode. The test set drives a current between the electrode under test (E) and a remote current probe (C2), then measures the voltage between E and a potential probe (P2) placed in a straight line between them. Resistance is voltage over current.
The catch is that every electrode has a "resistance area" — a volume of earth around it where most of the voltage drop happens. If the potential probe sits inside the resistance area of either the electrode or the current probe, the reading is wrong. Move the potential probe along the line and plot the readings, and you get a curve that climbs near the electrode, flattens out in the middle, then climbs again as P2 approaches C2. The flat section — the plateau — is the true resistance.
The 62% rule
For a uniform soil and a current probe far enough away, the plateau value falls at about 62% of the distance from the electrode to the current probe. That's why the field shortcut is to set P2 at 62% and take one reading. The shortcut only works if the probes are far enough apart — which is what the 52% and 72% readings prove.
The 52% / 72% plateau check
Take two more readings with P2 at 52% and 72% of the C2 distance. If both are within a few percent of the 62% reading, P2 is on the plateau and the 62% value is valid. The template calculates max deviation = the larger of |R52 − R62| and |R72 − R62|, divided by R62, compares it to your tolerance (5% by default — use the spec value if it gives one), and marks the run OK or NO PLATEAU.
No plateau means the current probe is too close. Move it farther out and repeat. For a single 10-foot rod, 100 feet usually works in open ground. For ring grounds, rod arrays and grids, the current probe needs to be much farther — the rule of thumb is at least 5 times the largest dimension of the electrode system, and many engineers use 5–10 times the grid diagonal. The template shows a minimum C2 distance based on the electrode dimension you enter.
Setting up the test in the field
- Isolate the electrode. Lock out the service and verify absence of voltage before lifting the grounding electrode conductor — or test before the utility ties in. With the GEC connected, the meter reads the electrode in parallel with every ground rod on the utility's multi-grounded neutral and you'll get a beautiful, meaningless number.
- Call 811 and look for buried metal. Water lines, gas lines, fence footings and rebar mats in the test path distort the readings. Run the probe line away from them.
- Drive the current probe at the distance you need, in a straight line. Record the distance and the direction.
- Take readings at 52%, 62% and 72%. The template calculates each probe distance from the C2 distance.
- Run a second direction at 90° or more from the first. Two runs that agree rule out buried-metal interference. The template reports the higher accepted reading — conservative, and hard to argue with.
- Record the soil and weather. Resistance depends heavily on moisture and temperature. A rod that reads 18 Ω after a wet week in April can read 35 Ω in a dry August or in frozen ground.
What resistance should you be aiming for?
| Target | Where it comes from |
|---|---|
| 25 Ω | NEC 250.53(A)(2) Exception — a single rod, pipe or plate electrode at 25 Ω or less doesn't need a supplemental electrode. It's a threshold for adding a second electrode, not a performance guarantee. Verify against the edition your AHJ enforces. |
| 5–10 Ω | Common in project specifications for industrial plants, telecom, data centers and critical facilities. |
| 1–5 Ω | Substations and large grids, set by the engineer's grid design. |
The spec wins. Enter the spec value as the target and the report returns PASS or FAIL with the margin. For the reasoning behind these numbers and what IEEE 81, IEEE 142 and IEEE 80 each cover, see our IEEE grounding standards explained reference.
When the reading is too high
- Add rods — the NEC minimum spacing is 6 feet, but rods spaced at least twice their length apart are far more effective because their resistance areas don't overlap.
- Go deeper with sectional rods; deeper soil is usually wetter and more stable year-round.
- Add a ground ring or use ground enhancement material around the electrode.
- Notify the engineer of record before you start driving rods on a spec'd system — they may want a specific design.
Retest after the fix and attach both reports. Sizing the grounding electrode conductor itself is a separate question — see the grounding electrode conductor chart.
The full profile — when one reading isn't enough
On grids, large ring grounds, or any result that's going to be disputed, take the full curve: readings every 10% of the C2 distance plus the 52/62/72% points. The template draws the curve in-cell so you can see the plateau (or the lack of one) at a glance. A curve with no flat section is the clearest possible evidence that the current probe needs to move.
The clamp-on (stakeless) method
Clamp-on testers induce a voltage in the ground lead and measure the resulting current around the loop. That loop is your electrode in series with everything else connected in parallel — so the method only works on multi-grounded systems: utility poles on a multi-grounded neutral, lightning protection down conductors, telecom tower legs. It is fast and needs no probes or disconnection, which is why it's popular for maintenance testing.
Its limits matter. It can't test a single isolated electrode. It reads very low (under about an ohm) when it finds a metallic loop instead of earth — a bonded fence or a second down conductor. And if there's measurable current on the ground lead, stop: that's a neutral or fault-path problem and a shock hazard. The Clamp-On tab records leakage current first, flags readings of 1 A or more, and flags suspiciously low readings for verification.
Related templates and references
- Switchgear & transformer startup checklist — attach the ground test to the pre-energization record
- Arc flash label data sheet — collect the data for the arc flash study
- Equipment grounding conductor chart
- Insulation resistance testing (IEEE 43)
- Inspection request log — schedule the ground inspection with the AHJ
- Browse all free construction templates
