Electrical
Ground ring electrode field guide (NEC 250.52(A)(4)) for electricians
What a ground ring is, the 20 ft of bare 2 AWG copper at 30 in deep, how to connect and bond it into the grounding electrode system, and why it gets done before backfill.
Direct answer
A ground ring is a loop of bare copper buried in the earth around a building and bonded into its grounding electrode system. NEC 250.52(A)(4) calls for at least 20 ft of bare copper no smaller than 2 AWG, encircling the structure, and 250.53(F) buries it at least 30 in deep. Confirm the figures against the adopted code.
Key takeaways
- NEC 250.52(A)(4) requires a ground ring of at least 20 ft of bare copper no smaller than 2 AWG, encircling the structure.
- NEC 250.53(F) buries the ground ring at least 30 in below grade, with no relief for hard digging; backfill to reach depth.
- Buried ring connections must be exothermic welds or connectors listed for direct burial under UL 467; unlisted clamps corrode underground.
- Per NEC 250.66(C), the GEC portion connecting to the ring need not be larger than the ring conductor itself.
- Inspect the ground ring before backfill; once the trench is closed the loop, depth, and welds cannot be verified without digging.
What a ground ring is and why big facilities use one
A ground ring is a loop of bare copper buried in the earth around a building, encircling the footprint, and bonded into the grounding electrode system. Because the copper follows the perimeter, every point along the structure has a short path down to the same earth connection, and the whole building sits on one low, even reference instead of a single point in the dirt at the service.
That is why it shows up on the big and the sensitive jobs. A driven rod talks to a narrow column of soil at one corner. A ring wraps the entire building in conductor and presses a long length of copper against a large volume of earth, so it reads low and holds that low reading through the dry season better than a rod does. For a data center, a tower, or a substation, the goal is not just to clear a fault. It is to hold every cabinet, every piece of structural steel, and every cable shield close to the same potential so lightning and switching surges do not push one part of the building up against another.
The ground ring is one of the electrode types the NEC recognizes, and it almost never works alone. It bonds together with the other electrodes present, the concrete-encased electrode, driven rods, metal water pipe, and building steel, into a single grounding electrode system. The overall system and the bonding rules are covered in the grounding electrode system and bonding guide, and the concrete-encased electrode has its own guide. This one stays on the ring.
What is a ground ring electrode?
A ground ring electrode is a length of bare copper conductor buried in a continuous loop that encircles the building or structure, in direct contact with the earth, and tied into the grounding electrode system. It is a perimeter electrode, not a point electrode. The copper is bare on purpose, so its entire surface is in contact with the soil and carries current to earth along the whole run, not just at the ends.
The loop is what sets it apart. A rod is a pin. A Ufer is a footing. A ring is a closed conductor that surrounds the footprint, so a fault or a surge anywhere on the building finds copper a short distance away in any direction. The ring also becomes the spine that other electrodes and structural steel land on, which is why on a large site it tends to be installed first and everything else gets bonded to it.
On the print it will be called a ground ring, a counterpoise, or a perimeter ground loop, and on tower and substation work the counterpoise name is common. They describe the same thing: a buried bare-copper loop serving as a grounding electrode.
What does NEC 250.52(A)(4) require for a ground ring?
NEC 250.52(A)(4) defines the ground ring as encircling the building or structure, in direct contact with the earth, and consisting of at least 20 ft of bare copper conductor not smaller than 2 AWG. Those are the three numbers to carry: at least 20 ft, bare copper, and 2 AWG minimum. Burn in the 20 ft of #2 bare copper, because that is the line between a ring electrode the inspector accepts and a piece of wire in a trench.
Read each word, because each one is a requirement. Encircling means the loop goes around the structure, not a partial arc along one side. Bare copper means no insulation and copper, not aluminum, for this electrode. And not smaller than 2 AWG is a floor, not a target. Designers routinely call out 4/0 copper or larger for the ring on substations and data centers, both to lower resistance and to survive fault and lightning current, so the spec on your job may be far heavier than the code minimum. The 2 AWG is the smallest the code will let you call a ground ring.
Confirm these figures against the edition your jurisdiction has adopted and any local amendments. The 20 ft length and the 2 AWG minimum have held across recent code cycles, but article and subsection numbers move, so verify before you cite them on a drawing or argue them with an inspector.
How deep does a ground ring have to be buried?
A ground ring is buried at least 30 in below the surface of the earth. That depth requirement sits in NEC 250.53(F), separate from the electrode definition in 250.52(A)(4), and it is the number people miss because they read the electrode spec and stop there. The ring has to encircle the structure and it has to be down at least 30 in. Both, not either.
The depth is about staying in soil that stays wet and stays put. The top foot or so of earth dries out, freezes, and swings with the weather, and an electrode up in that zone loses contact with moist soil exactly when you need it. Get the ring down below the frost line and into soil that holds moisture year round and it keeps a stable, low resistance through the dry months and the cold ones. The same logic is why a deeper rod usually beats a wider spread.
The 30 in is a minimum with no relief built in for hard digging. If you hit rock or an obstruction, the move is to bring fill back over the conductor to make the 30 in, route around the obstruction, or fall back to another electrode type, not to lay the ring shallow and hope. Verify the depth figure against the adopted code edition, since this is a section number and value worth confirming before you commit it to the field.
Why a buried loop reads low and even
Resistance to earth comes down to how much conductor surface is in contact with how much moist soil. A long run of bare copper, buried where the dirt stays damp, contacts an enormous volume of earth compared to a single rod, so the path from the copper to remote earth has little to push against and the reading comes in low.
The shape is the other half of it. Because the copper follows the perimeter, the resistance is low not just at one point but all the way around the building, and the potential in the soil under and around the structure is held even. That uniformity is the reason sensitive facilities want a ring. During a ground fault or a lightning strike, the earth around an electrode rises in potential, and if that rise is concentrated at one rod you get a steep voltage gradient across the ground that equipment, cable shields, and people can be caught between. Spread the electrode around the whole footprint and the rise is gentler and more uniform, so the difference in potential between any two points a person or a cabinet bridges stays small.
Adding length and adding rods on the loop both push the resistance lower, but the bigger payoff of the ring is the even reference. A low number on a single rod still leaves the rest of the building far from the electrode. The ring brings the earth connection to every wall.
Who actually installs a ground ring
Ground rings show up on the jobs where a single rod or a Ufer alone will not hold the building at a tight enough, even enough reference. Data centers are the headline case: rows of sensitive gear that share signal and power references cannot tolerate large potential differences across the floor, so the building gets a perimeter ring tied to the foundation steel and, inside, to a signal reference grid under the equipment.
Communication towers and broadcast sites use a ring, often called a counterpoise, because a tower is a lightning magnet and the strike energy has to be spread into the earth fast and wide. Substations are built on a buried ground grid that the perimeter ring is part of, sized heavy to carry fault current and to keep step and touch voltages within safe limits for anyone standing on the yard during a fault. Large commercial and industrial buildings, hospitals, and anything with a lightning protection system or a lot of sensitive electronics fall in the same bucket.
The common thread is low impedance for equipment and for lightning, around the whole structure, not just a fault path at the service. When the spec calls for a ring, it is usually because the design needs the even reference and the surge path, and that requirement comes from the engineer and the project documents, since the NEC requires a ring be used as an electrode only where one is already present at the building.
Ground ring vs ground rod vs Ufer
All three are recognized electrodes, and they differ in how much of the building they reference and how their resistance holds up. A driven rod is a point electrode: cheap, fast, and fine for an ordinary service, but it touches a narrow column of soil at one spot and its reading swings with the weather. A concrete-encased electrode, the Ufer, turns the footing steel into an electrode and reads low and stable because the concrete holds moisture against a large area, which is why on new construction it is the first electrode to use where the footing has qualifying steel. The concrete-encased electrode has its own guide.
The ring is the one that references the whole footprint. Where a rod and a Ufer give you a strong connection at a point or along the footing, the ring wraps the entire perimeter, so the low resistance and the even potential reach every side of the building. That is the property that matters for a data center, a tower, or a substation, and it is why those jobs specify a ring on top of, not instead of, the Ufer and the rods.
These are not competing choices on a big job. They are bonded together into one grounding electrode system, each doing what it does best, with the ring tying the perimeter together. How all the electrodes bond into one system, and the 25-ohm rod rule, is covered in the grounding electrode system and bonding guide.
| Electrode | What it references | Resistance behavior |
|---|---|---|
| Driven rod | One point at the service | Swings with weather, higher in dry or rocky soil |
| Concrete-encased (Ufer) | The footing it sits in | Low and stable, concrete holds moisture |
| Ground ring | The whole building perimeter | Low and even around the footprint |
What size GEC does a ground ring need?
The grounding electrode conductor that connects to a ground ring is not required to be larger than the ring conductor itself. That allowance is in NEC 250.66(C). If the ring is 2 AWG copper, the portion of the GEC running to the ring need not be larger than 2 AWG. If the design specified a 4/0 ring, that portion of the GEC need not be larger than 4/0. The logic matches the other electrode caps in 250.66: the conductor to an electrode does not have to be larger than what the electrode can use.
This parallels the caps you already know. The GEC to a rod, pipe, or plate need not be larger than 6 AWG copper, and the GEC to a concrete-encased electrode need not be larger than 4 AWG copper. The ring cap is different in that it is tied to the size of the ring you installed rather than a fixed number, because a ring can be anything from 2 AWG up to 4/0 or larger depending on the design.
Watch the shared run, the same as with any electrode. The main GEC sized off the service conductors from Table 250.66 can be larger than the ring tap, and a single GEC that serves the ring and other electrodes has to satisfy the largest requirement along its path. The 250.66(C) allowance only relaxes the portion that connects to the ring. Verify the section against the adopted code before you put it on a submittal. GEC sizing for the whole system is covered in the grounding electrode system and bonding guide.
Connections have to survive being buried
Every joint in a ground ring ends up underground, in wet soil, where you cannot get back to it, so the connection method is not a detail to leave to whatever is on the truck. The buried connections, ring to ring, ring to rod, and ring to the conductors that come off it, are made by exothermic welding, commonly the cadweld process, or by connectors listed for direct burial. A bare split bolt, a hardware-store clamp, or a plain zinc connector is a callback sealed in the dirt.
Exothermic welding is the standard on substation and tower rings because it fuses the copper into one continuous piece with no mechanical joint to loosen, corrode, or back off under fault and lightning current. Where a mechanical connector is used instead, it has to be listed for direct soil burial, marked accordingly, and rated for the conductors and the electrode it joins. The listing standard for grounding and bonding connectors is UL 467, and the buried ones carry the direct-burial rating on top of that.
Corrosion is the slow killer of buried connections. Dissimilar metals in wet earth set up a galvanic cell, and the less noble metal sacrifices itself until the joint goes high-resistance or opens, with nothing visible above grade to warn you. An exothermic weld removes the mechanical joint entirely, which is why it is the durable choice for the connection you will never see again. The connection you cannot reach is the one that has to be right the first time.
Bond the ring into the grounding electrode system
A ground ring is one electrode in the grounding electrode system, not a separate ground. Under NEC 250.50, every electrode present at the building is bonded together into one system, so the ring ties together with the concrete-encased electrode, the driven rods, the metal underground water pipe where it qualifies, and effectively grounded building steel. The aim is one common earth connection, not a ring at one potential and a Ufer at another.
That bonding is the safety point. Two grounded systems sitting at different potentials, with equipment or a person able to touch both, is a shock and damage path during a fault or a surge. Bonding the ring to the other electrodes removes the difference, so the whole installation rises and falls together. The bonding jumper that ties electrodes together is sized like a grounding electrode conductor, and the connection rules in 250.53(C) and 250.70 apply.
On a large job the ring is usually the piece everything lands on. The foundation steel ties to it, the rods on the loop tie to it, the service GEC lands on it, and the lightning protection system bonds to it. The full picture of which electrodes have to be present and how they all bond is in the grounding electrode system and bonding guide. For the ring specifically, the rule is plain: it is part of the system, so tie it in, do not let it stand alone.
Driving rods on the ring to pull the resistance down
A ring already reads low, but where the spec wants it lower, the common move is to drive ground rods at intervals around the loop and bond them to the ring. Each rod reaches down into deeper, often moister soil and adds a parallel path to earth, so the rods and the ring together read lower than the ring alone. On substations and tower sites you will see rods at the corners and at spacing around the perimeter, all welded to the loop.
Spacing matters, the same as it does for a pair of service rods. Rods set too close together share the same shell of soil, their fields overlap, and the second rod barely helps. A practical rule on a ring is to space the rods at least as far apart as they are deep, and farther where the design allows, so each rod works a fresh volume of earth. Depth usually beats crowding, because it reaches soil that stays wet year round.
The rods bond to the ring with the same buried-rated connections as the rest of the loop, exothermic or listed for direct burial. A rod welded to the ring is one continuous electrode. A rod near the ring with a corroded clamp is two separate problems waiting in the dirt.
The ring and lightning protection
On a structure with a lightning protection system, the ground ring is usually where the down conductors terminate into earth. The lightning protection standard, NFPA 780, calls for grounding the down conductors and for interconnecting the lightning protection grounding with the building's grounding electrode system, so the strike current has a wide, low-impedance path into the ground and the whole system rises together instead of one ground reference jumping against another.
The ring suits this job because a strike needs to be spread into the earth fast, and a perimeter loop with rods does that across the whole footprint rather than dumping the energy at one point. Tie each down conductor to the ring, bond the ring to the structural steel and the electrical grounding electrode system, and the lightning energy splits across a large volume of copper, steel, and soil. Concentrate it at a single electrode and you get a steep potential rise and side flashes to nearby metal.
Lightning protection design is its own discipline under NFPA 780, and the bonding between the lightning system and the electrical grounding electrode system is required, not optional, so coordinate the two rather than running them as separate grounds. Confirm the bonding details and the standard against the adopted edition and the project's lightning protection design.
Soil resistivity sets what you can achieve
How low a ring can read is set first by the soil, not by the copper. Soil resistivity, how hard the earth resists current, varies enormously with soil type, moisture, temperature, and the salts dissolved in it. Wet clay reads low. Dry sand, gravel, and rock read high, sometimes by orders of magnitude, and no amount of copper turns high-resistivity ground into low-resistivity ground on its own.
On a job where the resistance target is real, the soil gets measured before the design is finished, using a four-point soil resistivity test, the Wenner method. Four probes are driven in a line at equal spacing, a known current is passed through the outer pair, the voltage is read across the inner pair, and the resistivity is calculated from the spacing and the reading. Vary the spacing and you read resistivity at different depths, which tells the designer how deep the moist, low-resistivity soil is and how big the ring and how many rods it will take to hit the number.
This is the test that drives the design, and it is worth doing before the trench is dug, not after the ring is in and reading high. If the soil is dry and rocky, the answer is a larger ring, more and deeper rods, or ground-enhancing backfill around the conductor, and you want to know that before the excavation is open, not when the resistance test fails at the end of the job.
How do you test a ground ring's resistance?
The textbook field test for an electrode's resistance to earth is the fall-of-potential test, sometimes called the three-point test. You inject a current between the electrode under test and a remote current probe, then move a potential probe along the line between them and read the resistance, looking for the flat spot in the curve that represents the electrode's true resistance to remote earth. The method wants distance, often well over a hundred feet of clear run for the probes, and it gets awkward on a large ring tied into a whole building.
Know what the 25 ohm figure means before you quote it. The NEC's 25 ohm number applies to a single rod, pipe, or plate electrode: under 250.53, a single such electrode has to be supplemented with a second electrode unless it tests 25 ohms or less. It is a trigger for adding an electrode, not a performance target for a ring, and a ground ring is not a single rod. The code does not set a required resistance number for a finished ring.
What does set the target is the spec and the design. Sensitive facilities, substations, and tower sites routinely call for resistance values far below 25 ohms, single digits or lower, and those low numbers are spec values driven by IEEE guidance, not code mandates. IEEE 142, the Green Book, gives the low working targets for commercial and industrial systems, and IEEE 81 covers how the measurement is actually made. Measure to the spec, and read the result understanding it reflects the whole bonded system, not the ring in isolation.
Install the ring early, in the foundation trench
A ground ring is an excavation job, and the window to do it cheaply is short. The loop goes in the foundation trench, around the footprint, before the backfill goes back over it. Once the trench is closed and the site is graded and paved, putting in a ring means trenching all over again around a finished building, which is a different and far more expensive job.
So the ring gets coordinated with the excavation, the same way the Ufer connection gets coordinated with the concrete pour. Get on the schedule, know when the foundation trench will be open, and have the conductor, the rods, and the welds ready to go in that window. The connections to the foundation steel, the rods, and the stub-ups for the service GEC and the lightning down conductors all get made while the trench is open and the ring is still visible.
The failure mode is the familiar one on buried work: the site crew backfills on its own schedule, nobody told the electrician, and the trench is closed before the ring went in or before it was inspected. Now the electrode is either missing or buried unverified. Treat the backfill date as the deadline and the open trench as the hold point.
The ring goes in the dirt, not in conduit
A ground ring is an electrode, and an electrode works by being in direct contact with the earth. The bare copper goes in the soil, against the soil, for its whole length. Run it through conduit and you have insulated it from the very earth it is supposed to connect to, so it is no longer an electrode. The same goes for routing it above grade or laying it on top of a vapor barrier or a layer of gravel that separates it from moist soil.
This trips crews who are used to protecting conductors. Out of habit someone sleeves the ring through PVC where it crosses under a footing or comes up at a stub, and the sleeved section stops contacting earth. Short protective sleeves at a penetration are one thing, but the body of the ring has to be in the dirt. Bare copper, in direct earth contact, buried at depth, is the whole definition.
Where the conductor comes up out of the ground to land on a connection point or to bond to steel, that transition is the part you protect, not the buried loop. Keep the loop in the soil and keep it bare.
Inspect the ring and the welds before backfill
Once the trench is backfilled, the ring is gone from sight and the connections cannot be verified without digging. That makes the inspection a before-backfill event, not an after-the-fact one. The inspector who signs off on the ring has to see the loop in the trench, the depth, the welds and listed connections, and the bonding to the foundation steel and the other electrodes, while it is all still exposed.
Build it into the schedule like a footing or rebar inspection. Call for the grounding inspection while the trench is open, get it looked at, and get the sign-off before releasing the backfill crew. Photographs of the ring, the depth, and each connection before the dirt goes back are worth more than any note, because they are the only visual proof of a system that is about to disappear.
Pour the backfill first and the honest options are all bad: excavate to expose the ring, or argue that an unverified buried electrode is good. Neither is where you want to be when the resistance test at the end of the job reads high and nobody can see why. The connection you cannot inspect later is the one you inspect now.
Data centers: the ring and the signal reference grid
In a data center the ground ring is the outer layer of a larger grounding scheme. The perimeter ring bonds to the foundation steel and to a signal reference grid inside, the bonded copper grid run under the equipment that holds sensitive gear at a common potential across the whole floor. Tying the ring, the structural steel, and the signal reference grid into one system keeps cabinets, cable trays, and power references close to the same potential, which is what keeps surges and switching transients from appearing as a voltage difference between two pieces of equipment.
The detail to get right is the same as on any ring, with more connection points: plan the ring and the stub-ups before the trench closes and the slab is poured, use buried-rated connections, and coordinate with every trade that touches the slab and the floor. There are simply more places to bond and more chances to miss one. The signal reference grid and the equipment bonding inside the room are a topic of their own, beyond the ring.
Where ground rings go wrong
When a ring underperforms or fails inspection, the cause is almost always one of a short list, and they repeat across jobs because they are the steps that get cut when the trench has to close.
The conductor is too small or the loop too short, smaller than 2 AWG or not a continuous loop around the structure, so it does not meet the electrode definition. The ring is buried too shallow, up in the dry, freezing top layer instead of down at least 30 in in moist soil, so the resistance climbs in the dry season. A buried connection was made with a non-listed clamp or a mechanical joint instead of an exothermic weld or a direct-burial-listed connector, and it corrodes or loosens where nobody can see it. The ring was left as a separate ground, not bonded into the rest of the grounding electrode system, so it sits at its own potential. The ring went in late, or got backfilled before anyone inspected it, so it cannot be verified. And the loop was run in conduit or above grade instead of in direct earth contact, so it is not really an electrode at all.
Every one of these is cheap to prevent with the trench open and expensive to fix after the backfill. The list is the same one the inspector carries, so walk it yourself before you call for the inspection.
What to document
A ground ring disappears into the trench, so the record is the only proof of what got built. The point of the documentation is that the inspector, the next electrician, and the commissioning agent can each confirm the ring without excavating it, and that the buried connections can be defended years later when a resistance test or a fault investigation puts the grounding system in question.
Capture the ring conductor material and size, that it is bare copper and a continuous loop around the structure, the burial depth, the connection method and that the buried joints were listed or exothermic, where the stub-ups land, the GEC and bonding-jumper sizes, the soil resistivity and any resistance test result with the method used, and the date and sign-off of the before-backfill inspection. Photographs of the loop, the depth, and each weld before backfill belong in the record too, because they are the only visual proof once the trench is closed. Where the spec set a resistance target, record the measured value against it and verify the section numbers and figures against the adopted code edition.
| Field to record | Why it matters |
|---|---|
| Ring material, size, bare, continuous loop | Proves it meets the 250.52(A)(4) electrode spec |
| Burial depth | Confirms the 250.53(F) minimum depth |
| Connection method, listed or exothermic | Buried joints cannot be re-inspected later |
| Photos before backfill | Only visual proof once the trench is closed |
| GEC and bonding-jumper sizes | Shows 250.66(C) and bonding sizing were met |
| Soil resistivity and resistance test | Documents the design basis and the spec target |
| Before-backfill inspection date and sign-off | Ties the electrode to an approval |
Common mistakes
- Using a conductor smaller than 2 AWG or a loop shorter than 20 ft, or not actually encircling the structure.
- Burying the ring shallower than 30 in, up in the dry, freezing top layer instead of moist soil.
- Making a buried connection with a non-listed clamp or a mechanical joint instead of an exothermic weld or a direct-burial-listed connector.
- Leaving the ring as a separate ground instead of bonding it into the rest of the grounding electrode system.
- Installing the ring late or backfilling before the before-backfill inspection, so it cannot be verified.
- Running the loop in conduit or above grade instead of in direct earth contact, so it is not an electrode.
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.
Standards and references
The ground ring lives in NEC Article 250. The electrode itself is defined in 250.52(A)(4), which sets the encircling loop, the direct earth contact, the 20 ft minimum length, and the 2 AWG minimum bare copper. The burial depth of at least 30 in is in 250.53(F). The grounding electrode conductor to the ring is sized under 250.66, with the allowance in 250.66(C) that the portion connecting to the ring need not be larger than the ring conductor. Bonding the ring together with the other electrodes into one system is required by 250.50, the bonding jumper provisions are in 250.53(C), and electrode connections follow 250.70 with the listing standard UL 467 for the connectors, including the direct-burial rating for buried joints.
Article and subsection numbers move between code cycles, and the 20 ft length, the 2 AWG minimum, and the 30 in depth are the figures to confirm against the edition the jurisdiction has actually adopted and any local amendments. They have held across recent editions, but verify before you put them on a submittal or argue them with an inspector.
Beyond the NEC, IEEE 142, the Green Book, gives the low working resistance targets for commercial and industrial grounding, and IEEE 81 covers ground resistance and soil resistivity measurement. Lightning protection grounding and the required bonding to the electrical grounding electrode system come from NFPA 780. The AHJ has the final say on what is accepted and when it has to be seen. Stress the three things that decide whether this electrode succeeds: it must be at least 20 ft of bare 2 AWG copper encircling the structure, it must be in direct earth contact at least 30 in deep, and it must be bonded into the grounding electrode system and inspected before backfill.
Units and terms
The same loop goes by several names across a drawing set, a spec, and the field, so it helps to know the synonyms before they confuse a submittal.
Ground ring is the NEC term. Counterpoise is the tower and broadcast name, and perimeter ground loop shows up on commercial prints. Conductor size is given in AWG for the smaller sizes and kcmil for the larger ones, with 2 AWG the code minimum and 4/0 or larger common on heavy designs. Depth is in inches, and the minimum is 30 in. Resistance to earth is in ohms, and soil resistivity is in ohm-meters or ohm-centimeters. GEC is the grounding electrode conductor, the wire from the electrode to the service, and it is easy to confuse by name with the EGC, the equipment grounding conductor, which carries fault current and is a different thing.
- Ground ring / counterpoise
- A buried bare-copper loop encircling a structure, used as a grounding electrode under NEC 250.52(A)(4)
- 2 AWG
- The minimum bare copper conductor size for a ground ring, with heavier sizes common by design
- GEC
- Grounding electrode conductor, the wire connecting the electrode to the grounded service equipment
- Exothermic weld
- A permanent molecular bond from a controlled reaction, the durable choice for buried ring connections
- Soil resistivity
- How strongly the earth resists current, measured by the Wenner four-point test, set in ohm-meters
- Fall-of-potential
- A ground-resistance test that injects current and reads the flat spot in the potential curve
- Signal reference grid
- A bonded copper grid under data center equipment that holds the floor at a common potential
FAQ
What is a ground ring?
A ground ring is a loop of bare copper buried in the earth around a building, encircling the footprint, and bonded into the grounding electrode system. Because it follows the perimeter, every part of the structure has a short path to earth at a low, even resistance, which is why data centers, towers, and substations use one.
What size wire is required for a ground ring?
NEC 250.52(A)(4) requires bare copper conductor not smaller than 2 AWG, in a continuous loop encircling the structure. That 2 AWG is a minimum, not a target. Substation and data center designs often specify 4/0 copper or larger to lower resistance and carry fault and lightning current. Confirm the size against the adopted code and the project spec.
How deep is a ground ring buried?
A ground ring is buried at least 30 in below the surface of the earth, under NEC 250.53(F). The depth keeps the conductor in moist soil below the dry, freezing top layer, so the resistance stays low and stable. There is no relief for hard digging, so backfill to reach 30 in rather than laying it shallow.
When do you need a ground ring?
A ground ring is used where the design needs a low, even earth reference around the whole structure: data centers, communication towers, substations, and large or sensitive buildings, often with lightning protection. The NEC requires using a ring only where one is already present, so the requirement to install one usually comes from the engineer and the project documents.
How do you connect a ground ring underground?
Buried ground ring connections are made by exothermic welding, often the cadweld process, or with connectors listed for direct burial under UL 467. A welded joint fuses the copper into one piece with nothing to loosen or corrode. Unlisted clamps and mechanical joints corrode in wet soil where you cannot reach them, so they are not acceptable buried.
What size GEC connects to a ground ring?
Under NEC 250.66(C), the portion of the grounding electrode conductor that connects to a ground ring need not be larger than the ring conductor itself. A 2 AWG ring takes a 2 AWG tap, a 4/0 ring a 4/0 tap. A shared GEC serving several electrodes still has to meet the largest requirement along its run.
Does a ground ring have to meet a resistance value?
The NEC does not set a required resistance for a finished ground ring. The 25 ohm figure applies to a single rod, pipe, or plate as a supplemental-electrode trigger, not to a ring. Low targets like single-digit ohms come from IEEE 142 and the project spec, not the code, so measure to the spec and verify the adopted edition.
Can a ground ring be installed in conduit?
No. A ground ring is an electrode and works by direct contact with the earth, so the bare copper goes in the soil for its whole length. Running it in conduit or above grade insulates it from the earth, so it is no longer an electrode. Protect only the transition where it surfaces, not the buried loop.
When does a ground ring get inspected?
Before backfill. Once the trench is closed, the ring and its connections are buried and cannot be verified without digging. The inspector has to see the loop, the depth, the welds, and the bonding while they are still exposed, so the grounding inspection rides in the same window as the open trench. Photograph each connection before the dirt goes back.
What is the difference between a ground ring and a ground rod?
A ground rod is a point electrode touching a narrow column of soil at one spot, and its resistance swings with the weather. A ground ring is a buried loop around the whole building, so it reads low and holds an even potential around the footprint. On big jobs the two are bonded together, not chosen one over the other.
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.