Datacenter
Ground resistance and bonding testing field guide for data centers
Test ground resistance with fall-of-potential, prove the bonds with a ductor, and record the soil conditions that drove the number.
Direct answer
Ground resistance testing measures the resistance from a grounding electrode system to remote earth, in ohms, usually by the fall-of-potential method. Bonding testing verifies low-resistance connections between metal parts. NEC requires a supplemental electrode unless a single rod tests 25 ohms or less; lower targets like 5 ohms are spec values, not code.
Key takeaways
- NEC 250.53(A)(2) requires a supplemental electrode unless a single rod tests 25 ohms or less; 25 ohms is a trigger, not a target.
- IEEE 142 (Green Book) cites 1 to 5 ohms for large commercial and industrial systems, and 1 ohm or less for sensitive sites.
- The 62 percent rule places the potential probe at 61.8 percent of the electrode-to-current-probe distance, valid only in uniform soil.
- Ground resistance is measured in ohms by fall-of-potential; bonding is measured in milliohms with a low-resistance ohmmeter (ductor/DLRO).
- A stakeless clamp tester works only on multi-grounded systems with a parallel return path; it cannot read an isolated single electrode.
Ground resistance testing and bonding testing, defined
Ground resistance testing measures the resistance from a grounding electrode system to remote earth, expressed in ohms. The current that flows into a fault, a lightning strike, or a surge has to spread out into the soil through your electrodes, and the soil fights that spread. That opposition is what the test reads. It is not the resistance of a wire. It is the resistance of dirt around metal.
Bonding testing is a different measurement that gets confused with it constantly. Bonding verifies that two metal parts are tied together with a low-resistance connection, measured in milliohms or fractions of an ohm with a low-resistance ohmmeter. Ground resistance is the path to earth. Bonding is the path between parts. A data center grounding job needs both proven, and the two use different instruments and different acceptance numbers.
Here is what a ground resistance test does not tell you. It does not prove your bonds are tight. It does not prove the electrode is connected to anything. A buried rod with a broken clamp can still read a fine resistance to earth on a stakeless clamp and be doing nothing for the system it is supposed to protect. You test the resistance and you test the bonds, separately, because each one fails in its own way.
What a low ground actually buys you
Grounding does several jobs, and they do not all want the same thing, which is why the targets argue with each other.
Fault clearing comes first, and it is mostly about bonding, not earth resistance. When a hot conductor faults to a metal enclosure, the fault current returns through the equipment grounding conductor back to the source, not through the dirt. That low-impedance bonded path is what trips the breaker fast. The earth electrode does almost nothing for a line-to-ground fault on a grounded system. People get this backward and chase a low earth reading thinking it clears faults. It does not. The bond does.
Where earth resistance earns its keep is lightning, surge, and reference. A low-impedance path to earth gives a lightning strike and a surge protective device somewhere to dump energy, and the lower and shorter that path, the less the voltage rises on everything bonded to it. Sensitive electronics want a stable zero-volt reference, and a high-resistance ground lets that reference float and bounce. And step and touch potential, the voltage a person can bridge during a fault, is a soil-and-geometry problem that a good grounding grid controls and a single high-resistance rod does not.
How does the fall-of-potential (3-point) test work?
The fall-of-potential method, also called the 3-point method, drives a known current between the electrode under test and a remote current probe, then reads the voltage between the electrode and a movable potential probe to find the resistance. It is the reference method in IEEE 81 and the one an acceptance spec usually means when it says test the ground.
You set three things in a line: the electrode under test, a potential probe in the middle, and a current probe far out. The tester pushes current between the electrode and the current probe and measures voltage between the electrode and the potential probe. The trick is that the electrode and the current probe each have a resistance shell around them in the soil, and if the potential probe sits inside either shell, the reading is wrong.
The 62 percent rule places the potential probe at 61.8 percent of the distance from the electrode to the current probe, rounded to 62 percent, where the two shells cancel and the reading reflects true resistance to earth. That number falls out of the math for a single electrode in uniform soil. It is not a guess and it is not adjustable. Get the current probe far enough out, put the potential probe at 62 percent, and you are reading earth.
Getting a flat plateau, and why probe spacing decides the test
One reading at 62 percent is a number. A plateau is proof. Take three readings with the potential probe at 52, 62, and 72 percent of the electrode-to-current-probe distance, plot them, and if they sit on a nearly flat line the test is valid. If the three readings climb steeply, the current probe is too close and the resistance shells overlap, so you move the current probe farther out and run it again.
How far is far enough depends on the size of what you are testing. A single rod needs the current probe out maybe 100 to 130 ft to get clear of its shell. A large grounding grid or a ring around a data center has a resistance area that can extend hundreds of feet, and the current-probe distance grows with it, sometimes to several times the diagonal of the grid. Test a big grid with short leads and you will read a low number that is a lie.
The 62 percent rule assumes uniform soil. Layered soil, common on real sites, shifts where the flat portion sits, anywhere from 10 to 90 percent of the span, and you cannot find it by eye. That is when you stop trusting the single rule and run the full traverse, plotting the whole curve to find where it actually flattens.
Field example: a fall-of-potential traverse on a ground ring
A ground ring around a small data center module reads through a fall-of-potential traverse with the current probe set 300 ft out. We walk the potential probe out in steps and record resistance at each, looking for the flat middle.
The readings climb fast near the ring, flatten across the middle third, then climb again as the probe nears the current electrode. The flat band sits around the 62 percent point, 186 ft, and the three readings at 52, 62, and 72 percent agree within a few hundredths of an ohm. That agreement is the pass. We record the plateau value, not the average of the whole walk, because the ends of the walk are inside the resistance shells and do not represent earth.
If the middle had not flattened, the fix is not to average the mess. It is to push the current probe from 300 ft to 500 ft or more and walk it again. A traverse that never flattens is a traverse that has not cleared the resistance areas, and no amount of arithmetic rescues it.
| Potential probe distance | Percent of span | Resistance read |
|---|---|---|
| 90 ft | 30 percent | 2.1 ohms |
| 156 ft | 52 percent | 3.3 ohms |
| 186 ft | 62 percent | 3.4 ohms |
| 216 ft | 72 percent | 3.5 ohms |
| 270 ft | 90 percent | 4.9 ohms |
What is the four-point (Wenner) soil resistivity test for?
The four-point test, the Wenner method, measures soil resistivity to design a grounding system, not to test one that is already in. This is the distinction people miss. Fall-of-potential tells you what an installed electrode reads. Wenner tells you what the dirt will let you achieve before you drive anything.
You set four equally spaced pins in a straight line, spacing a between each. The outer two inject current, the inner two read voltage, and the tester returns a resistance R. Soil resistivity is then 2 pi times a times R, with a in a consistent length unit, giving resistivity in ohm-meters or ohm-centimeters. The depth of soil that reading represents is roughly equal to the pin spacing, so you repeat the traverse at increasing spacings, commonly starting near 5 to 10 ft and stepping up by a factor around 1.5, to map resistivity with depth.
Why bother before driving rods? Because resistivity decides everything. It tells the designer whether a 10 ft rod lands in conductive clay or useless dry sand, whether to go deeper, spread wider, or build a ring, and it feeds the substation grid math in IEEE 80. On a data center ground grid, skipping the resistivity survey means designing the electrode system blind and finding out at acceptance testing that the soil never supported the number.
ρ = 2 π a R- ρ (rho)
- Soil resistivity, in ohm-meters or ohm-centimeters
- a
- Equal spacing between adjacent pins, in a consistent length unit
- R
- Resistance returned by the four-point tester, in ohms
When does a clamp-on (stakeless) ground tester work?
A clamp-on, or stakeless, tester reads ground resistance without driving any probes, by clamping around the conductor and using the rest of the grounded system as the return path. It is fast, it works on a live system, and it is the right tool for one specific situation: a multi-grounded system where many electrodes sit in parallel and the test current has a loop to flow around.
It cannot test an isolated electrode. No parallel return path means no loop, and the clamp reads open or garbage. Drive a single rod for a remote antenna, clamp it, and the number is meaningless. That is fall-of-potential work.
The trap that bites people is the false low reading. The clamp measures the electrode under test in series with the parallel combination of every other electrode in the loop. On a system with many rods, that parallel combination is small, so the clamp reads close to the one rod, which is fine. But if another electrode sits inside the resistance area of the one you are testing, the reading comes back lower than the rod's true resistance to earth, and you walk away thinking the ground is better than it is. A clamp number that looks too good on a big system deserves a fall-of-potential check before you certify it.
Selective testing: one electrode without disconnecting it
Selective testing is fall-of-potential with a clamp added so you can measure one electrode in a connected system without disconnecting it. You set the current and potential probes the same way as a standard fall-of-potential test, but you clamp the current transformer around the specific downlead you care about, so the tester reads only the current through that electrode and ignores the rest of the parallel network.
This is the method for a working facility where you cannot open the grounding system to isolate a rod. Disconnecting a ground to test it, on a live data center, is exactly the wrong move, because for the seconds it is open that part of the system has no ground. Selective testing keeps everything bonded and still gives you a true per-electrode number.
The limits are the probe limits of fall-of-potential, since that is what it is underneath. You still need room to set the current probe far enough out to clear the resistance area, and you still walk a plateau to trust the reading. The clamp solves the disconnection problem. It does not solve the probe-spacing problem.
Two-point and three-point variations, and the dead-earth limit
The full fall-of-potential test uses three or four terminals and a proper probe traverse. The shortcuts trade accuracy for speed, and each one has a hole you need to know about.
The two-point, or dead-earth, method measures the electrode under test in series with a second, assumed-good ground, often a metal water main, and reports the sum. If the reference ground is genuinely low, say a large municipal water system, the sum is close enough to use as a rough check. The hole is obvious. You are trusting a ground you did not measure, and on a plastic-pipe site or an isolated reference there is no good dead earth to lean on. The number is only as good as the assumption under it.
Three-terminal and four-terminal versions of the instrument exist mainly to deal with lead resistance. The four-terminal hookup cancels the resistance of the test leads, which matters when you are reading a low-resistance grid where a fraction of an ohm in the leads would skew the result. For a single rod reading several ohms, the three-terminal version is fine. For a sub-ohm grid, use the four-terminal connection so you are reading the ground and not your own wires.
What ground resistance is acceptable?
There is no single acceptable number, and treating one as universal is the most common mistake in this whole subject. The NEC sets a threshold, not a target. Under 250.53(A)(2), a single rod, pipe, or plate electrode has to be supplemented with a second electrode unless it is shown to have a resistance to earth of 25 ohms or less. That 25 ohms is a trigger for adding a rod, not a performance goal, and once you have two electrodes the NEC stops asking for a resistance number at all.
The low numbers everyone quotes come from somewhere else. IEEE 142, the Green Book, finds resistances in the 1 to 5 ohm range suitable for large commercial and industrial systems, with 5 ohms a common working target and 1 ohm or less cited for generating and transmission stations and for sensitive installations. Those are recommendations and findings, not code mandates. A telecom or data center spec calling for 5 ohms, or 1 ohm at the signal reference grid, is a contract value, and the contract is what you test against.
So the honest answer on a job is this: hit the number in the project specification. If the spec is silent, the NEC 25 ohm rule governs the supplemental-electrode decision, and good practice on a data center aims far lower because of lightning, surge, and reference, not because a code section forces it.
| Reference | Number | What it actually is |
|---|---|---|
| NEC 250.53(A)(2) | 25 ohms or less, single rod | Threshold to skip a supplemental electrode, not a target |
| IEEE 142 (Green Book) | 1 to 5 ohms | Recommended range for large commercial and industrial, not a mandate |
| IEEE 142, station/sensitive | 1 ohm or less | Recommendation for stations and sensitive sites |
| Data center / telecom spec | Often 5 ohms, 1 ohm at the SRG | Contract value, test against the spec |
Bonding and continuity testing with a low-resistance ohmmeter
Bonding testing proves that two metal parts are tied together well enough to carry fault current and equalize potential, and it is a milliohm measurement, not an ohm measurement. The instrument is a low-resistance ohmmeter, the ductor or DLRO, which pushes a known test current, often 10 A and up, through the connection and reads the voltage drop across it with a four-wire Kelvin hookup so the test leads do not corrupt the reading.
A regular multimeter will not do this. Its tiny test current reads through a corroded or loose bond as if it were fine, because milliamps find a path that fault amps cannot. The ductor's high current is what exposes the bad joint. A bond that reads a few milliohms is solid. One that reads in the ohms, or jumps around when you wiggle the lug, is a connection that will heat, arc, or open under real current.
In a data center you ductor the meaningful bonds: the grounding electrode conductor from the system to the electrode, the equipment grounding conductors and busbars, the connections between the signal reference grid members, and the bonds from racks, cable tray, raceway, and the raised floor to the common bonding network. Acceptance specs and NETA give pass values in milliohms or as a maximum drop. The principle is the same everywhere. Low and stable passes. High or jumpy fails.
The signal reference grid and the common bonding network
A data center bonds everything metal into one equipotential plane so that no two points the equipment touches can sit at different voltages. The signal reference grid, the SRG, is the copper mesh under or around the equipment that ties it all to a common high-frequency reference. The broader common bonding network, the CBN, is everything bonded together: rack frames, cable tray, conduit, raised-floor pedestals, busbars, and the SRG itself.
The myth to kill is the isolated ground. There is a persistent belief that sensitive equipment wants its own private, separated earth, kept away from the dirty building ground, to keep noise out. It is wrong and it is against code. A truly isolated ground creates a voltage difference between the equipment and everything around it, which is the hazard and the noise path you were trying to avoid, and the NEC requires a signal reference structure to be bonded to the equipment grounding conductor, not floated off on its own. The signal reference grid earned its reputation in the era of unbalanced signaling like RS-232 and coax. With balanced Ethernet and fiber, its noise role has shrunk, but the bonding-and-equipotential job remains.
What you test on the SRG and CBN is bonding, not earth resistance. You ductor the grid connections and the bonds to it, looking for the low, consistent milliohm readings that prove the plane is actually one plane and not a dozen islands that happen to be near each other.
Step and touch potential at the data center
Step and touch potential is the voltage a person can be exposed to during a ground fault, and it is the safety reason the grounding grid geometry matters as much as the resistance number. Touch potential is the difference between a hand on a grounded metal object and the feet standing nearby. Step potential is the difference between two feet a stride apart on the soil during a fault.
A low total earth resistance does not, by itself, make a site safe to stand on. You can have a low-ohm ground and still have dangerous gradients in the soil right around an electrode where the current is dumping. The control is a grid, a buried mesh that ties the surface to a common potential so there is little difference to bridge, sometimes with a crushed-stone surface layer that raises the resistance between a person's feet and the earth. IEEE 80 is the substation reference for designing and checking these gradients, and large data center yards with their own medium-voltage gear borrow directly from it.
The field point is that resistance testing and gradient safety are related but separate questions. You measure the ohms to know the system performs. You design and verify the grid geometry to know a person near a fault survives it.
Seasonal, moisture, and temperature effects on readings
Soil resistivity, and so your ground resistance reading, moves with water and temperature, and the swing is large enough to flip a pass into a fail. Wet soil conducts. Dry soil does not. The same rod that reads 4 ohms after a wet spring can read 8 or 10 ohms in a dry August, because the moisture that carried the current has left the top few feet of soil.
Temperature works the same direction at the cold end. As soil approaches and passes freezing, its resistivity climbs steeply, so a reading taken in frozen ground can be far worse than the summer number and is not representative of the system's normal condition. This is why a single test on a convenient day is a snapshot, not the truth.
The practical moves are two. First, record the weather and recent rainfall with every reading, because a number without its conditions cannot be compared to the next one. Second, where it matters, drive electrodes deep enough to reach soil that stays moist year round, below the zone that dries out and freezes, so the system's resistance is stable instead of seasonal. An acceptance test passed in ideal spring conditions on shallow rods can quietly fail in the dry season when nobody is testing.
What do I do if my ground resistance is too high?
When a reading comes back over the spec, you have a handful of fixes, and they do not all pay off equally. Rank them by what the soil is doing.
Drive deeper before you drive wider. A longer rod reaches into soil that is usually moister and more conductive with depth, and deeper rods hold their reading through the dry season, so a single deep rod often beats two shallow ones. Where one rod is not enough, add rods, but space them right. Two rods closer together than about one rod length share the same resistance area and their fields overlap, so you get far less than half the resistance. Space them at least their driven depth apart, farther is better, and the parallel benefit actually shows up.
Then the diminishing returns. The first added rod helps a lot. The second helps less. By the time you are paralleling many rods you are spending money for tenths of an ohm, and the better move shifts to a different geometry: a ground ring encircling the building, a concrete-encased electrode in the footings, or chemical and enhancement electrodes that condition the soil around them with salts to lower local resistivity. On genuinely bad soil, sand or rock, no number of plain rods gets you there, and the ring plus the concrete-encased electrode plus soil enhancement is the realistic path. Run the soil resistivity survey first so you are solving the soil you actually have.
The grounding electrode system components
The grounding electrode system is not one thing. It is every qualifying electrode at a building bonded together, and the NEC requires the ones present to be used and tied into a single system. Each type behaves differently in a test and in service.
Driven rods are the default, cheap and quick, and their resistance depends heavily on depth and soil moisture, which is why they swing with the season. A ground ring, a bare conductor buried in a loop around the structure, spreads the contact with earth over a large area and gives a stable, low reading, which is why data centers favor it. The concrete-encased electrode, the Ufer, is rebar or wire embedded in the footing concrete, and it is one of the better electrodes there is, because concrete holds moisture and contacts a large soil volume; it reads low and stays low. Structural building steel, where it qualifies, ties the whole frame into the system. A metal underground water pipe can serve as an electrode where it meets the code's contact-length and continuity rules, but it has to be supplemented, because plastic repairs and dielectric unions break the path silently.
When you test, you are usually reading the whole bonded system, not one piece. To check a single electrode you isolate it or use selective testing, and on a finished system that means the clamp, not a wrench on the bonding jumper.
The instruments, and using them right
Four instruments cover this work, and using the wrong one is how good grounds get certified bad and bad grounds get certified good.
The earth ground tester, three or four terminal, does fall-of-potential, soil resistivity, and selective testing with the right leads and clamp. It is the instrument an acceptance spec assumes. The stakeless clamp tester does fast in-service checks on multi-grounded systems and nothing useful on an isolated electrode. The low-resistance ohmmeter, the ductor or DLRO, does bonding and continuity in milliohms with its four-wire connection and high test current. A plain multimeter does none of these jobs correctly and should not appear in a grounding acceptance report.
Two habits separate a clean test from a questionable one. Use the four-terminal connection whenever you are reading low resistance, so the test leads cancel out instead of adding into the number. And keep the test leads, probes, and clamps in calibration with a documented date, because an acceptance test is only worth the traceability behind it. NETA acceptance testing expects calibrated instruments and recorded conditions, and an inspector or commissioning agent will ask for both.
Field checklist
Want this checklist to run itself on every job — with photo proof and a signed record crews can hand the customer? That's FieldOS.
What to document
A grounding test report that cannot be reproduced is not a report, it is a number on a page. The point of acceptance documentation is that someone can come back in five years, read it, and know whether the system has drifted. That means recording the conditions and the method, not just the ohms.
Capture the electrode or bond tested and its location, the method and instrument used, the probe spacings for a fall-of-potential test, the reading, the soil and weather, and the pass or fail against the specific spec value. When you record a bonding result, record it in milliohms with the test current the ductor used. Tie every result to the spec it was judged against, so the next person is not guessing which target applied.
| Field to record | Why it matters |
|---|---|
| Electrode or bond, and location | Identifies exactly what was tested |
| Method (fall-of-potential, clamp, selective, ductor) | Sets how the number should be read |
| Probe spacing and current-probe distance | Lets a reviewer judge if the plateau was valid |
| Reading (ohms, or milliohms for bonds) | The result itself |
| Soil condition, rainfall, temperature | Readings move with moisture and frost |
| Instrument and calibration date | Traceability the inspector will ask for |
| Pass or fail vs the spec value | Ties the result to the number that governs |
Common mistakes
- Clamping a stakeless tester on an isolated single electrode, which has no return path and cannot be measured that way.
- Setting the probes too close, so the resistance areas overlap and the reading comes back low and false.
- Accepting one reading instead of walking a 52, 62, 72 percent plateau to prove it.
- Reading a large grid with short leads, which never clears the resistance area and understates the resistance badly.
- Treating the 25 ohm NEC figure as a performance target instead of a supplemental-electrode threshold.
- Trusting a clamp-on low reading on a big parallel system without a fall-of-potential confirmation.
- Using a multimeter for bonding, whose low current passes a corroded joint that fault current would not.
- Disconnecting a working ground to test it instead of using selective testing, leaving the system ungrounded while open.
- Testing in ideal wet conditions and reporting it as the system's resistance year round.
Standards and references
The framework for this work lives in a few documents, and naming the right one for the right point is how a report holds up. IEEE 81 is the guide for measuring earth resistivity, ground impedance, and earth surface potentials, and it is the reference behind fall-of-potential, the 62 percent rule, and the Wenner traverse. IEEE 142, the Green Book, covers grounding of industrial and commercial power systems and is the source of the 1 to 5 ohm working figures, which it presents as recommendations, not requirements. IEEE 80 is the substation grounding safety guide that governs grid design and step and touch potential, and large data center yards lean on it.
On the code side, the NEC, NFPA 70 Article 250, governs the grounding and bonding installation. The single-rod supplemental-electrode rule and its 25 ohm threshold sit at 250.53, and the rod installation requirements for depth and spacing sit nearby in the same part, with supplemental electrodes generally spaced at least 6 ft apart. The bonding of a signal reference structure to the equipment grounding conductor is addressed in the information technology equipment provisions. Section numbers move between code cycles, so confirm them against the adopted edition and local amendments before citing them on a submittal.
NETA acceptance testing specifications give the acceptance procedures and pass criteria a commissioning agent tests to, and the manufacturer's instructions and the project specification override the general targets wherever they are stricter. Cite the standard that controls the point, and let the contract value win over the rule of thumb.
Units, terms, and conversions
Grounding work mixes a few unit systems, and a spec, an instrument, and a soil report can each use a different one for the same idea.
Ground resistance is in ohms. Bonding and continuity are small, so they read in milliohms, where 1000 milliohms is 1 ohm, and a good bond is a few milliohms. Soil resistivity is the property of the dirt itself, reported in ohm-meters in most engineering work and in ohm-centimeters in some North American sources, where 1 ohm-meter equals 100 ohm-centimeters. Keep the resistivity unit straight, because a factor of 100 error here sizes a grounding system wrong.
- Ground resistance / earth resistance
- Resistance from the electrode system to remote earth, in ohms
- Bonding resistance
- Resistance of the connection between two metal parts, in milliohms
- Soil resistivity
- The soil's own resistance property, in ohm-meters or ohm-centimeters
- Fall-of-potential
- The 3-point method that drives current and walks a potential probe to find true earth resistance
- 62 percent rule
- Potential probe at 61.8 percent of the electrode-to-current-probe distance, for a valid reading in uniform soil
- SRG / CBN
- Signal reference grid and common bonding network, the bonded equipotential plane in a data center
- GEC / EGC
- Grounding electrode conductor to the earth electrodes, and equipment grounding conductor for the fault path
FAQ
What is the 62 percent rule in fall-of-potential testing?
The 62 percent rule places the potential probe at 61.8 percent of the distance from the electrode to the current probe, rounded to 62 percent, where the resistance shells around each electrode cancel and the reading reflects true earth resistance. It holds in uniform soil and shifts in layered soil.
Is 25 ohms a good ground resistance?
25 ohms is not a target, it is the NEC threshold under 250.53(A)(2) for skipping a supplemental electrode on a single rod. Data center and IEEE practice aims far lower, commonly 5 ohms or 1 ohm, for lightning, surge, and reference. Test against the project specification, not the 25 ohm figure.
Clamp-on or fall-of-potential: which ground test should I use?
Use fall-of-potential to measure a single or isolated electrode accurately and for acceptance testing. Use a clamp-on stakeless tester for fast in-service checks on multi-grounded systems with a parallel return path. A clamp cannot read an isolated electrode and can show a falsely low number on large parallel systems.
What do I do if my ground resistance is too high?
Drive the rod deeper before adding more rods, since deeper soil stays moist and conductive. If you add rods, space them at least their driven depth apart so their fields do not overlap. On bad soil, switch to a ground ring, a concrete-encased electrode, or chemical electrodes.
What is the difference between ground resistance testing and bonding testing?
Ground resistance testing measures the path from the electrode system to earth, in ohms, with a fall-of-potential or clamp tester. Bonding testing measures the connection between two metal parts, in milliohms, with a low-resistance ohmmeter. One checks the path to dirt, the other checks the path between parts.
What is the four-point Wenner test used for?
The four-point Wenner test measures soil resistivity, in ohm-meters, to design a grounding system before it is installed. It uses four equally spaced pins and the formula resistivity equals 2 pi times spacing times measured resistance. It does not test an installed electrode; fall-of-potential does that.
How far should the current probe be in a fall-of-potential test?
Far enough to clear the electrode's resistance area, which depends on its size. A single rod may need 100 to 130 ft; a large grounding grid can need several times its diagonal, hundreds of feet. If the 52, 62, 72 percent readings do not flatten, move the current probe out and retest.
Can I test a single ground rod with a clamp-on meter?
No. A stakeless clamp-on tester needs a parallel return path through other grounded electrodes to read at all, so an isolated single rod gives a meaningless number. Use the fall-of-potential method with driven probes for a standalone rod, or use selective testing if the rod is part of a connected system.
Does a low ground resistance clear electrical faults?
No. On a grounded system a line-to-ground fault returns through the bonded equipment grounding conductor to the source, not through the soil, and that low-impedance bond is what trips the breaker. Earth resistance matters for lightning, surge, and reference, not for clearing line-to-ground faults. Test and prove the bonds separately.
People also ask
Codes cited in this guide
This guide is written and reviewed against the published standards below. Always confirm the current adopted edition with the authority having jurisdiction.