Electrical
Concrete-encased underground duct bank field guide for electrical crews
Lay out the conduit grid, hold the spacing and cover, tie it down so it does not float, and get the pre-pour inspection right because the concrete is the last word.
Direct answer
A concrete-encased duct bank routes feeder conduits underground in a grid of PVC encased in concrete, protecting the cable and letting many circuits share one trench. Spacing, cover, encasement, and heat all have to be right before the pour, because the concrete cannot be changed once it sets. The engineered design and project spec govern.
Key takeaways
- Minimum cover comes from NEC Table 300.5 for circuits up to 1000 V, measured to the top of the concrete; under vehicular traffic the table commonly calls for 24 in, and medium-voltage banks or specs often run 30 to 36 in.
- Duct-bank ampacity is lower because conductors heat each other; size from engineered figures (NEC Annex B for low voltage, 310.60 tables for 2001 to 35000 V, built on Neher-McGrath), not Table 310.16.
- The center conductor in a tightly packed bank can derate toward 60 percent of its free ampacity, so the detail spacing (commonly a 3 in clear minimum) is not negotiable.
- PVC ducts float in wet concrete like corks; tie the grid to staked rebar, pour in lifts, and avoid over-vibration to keep spacing and cover from blowing out.
- Mandrel-proof every duct before cable and pitch the bank to drain toward manholes (commonly about 3 in per 100 ft); a failed mandrel can reject the whole run, and a belly holds water against the cable for life.
Concrete-encased duct bank, and why you cannot fix it after the pour
A concrete-encased duct bank is a grid of conduits, run underground and surrounded on all sides by concrete, that carries feeders from one point to another while protecting the cable inside. Think of it as a permanent multi-circuit highway poured into the ground. The conduits hold the cable, the concrete protects the conduits and spreads the load, and the spacing between conduits controls how the circuits heat each other.
You build a duct bank when many circuits go the same way, when the run crosses under a road or a yard that will see traffic, or when the feeders are valuable enough that direct burial is too much risk. A data center campus is the usual case: dozens of medium-voltage and low-voltage feeders leaving a substation or utility yard, all headed for the same buildings, all needing protection and a known thermal environment.
The hard part is not the digging. It is that the concrete is the last word. Spacing wrong, cover short, a conduit crushed, a duct not proofed, no slope to drain, and you find out after the pour, when the fix is a jackhammer and a re-pour instead of a tie-wire. Everything in this guide happens before the concrete truck shows up, because that is the only time any of it is cheap to fix.
Conduit type and the duct bank layout
The common duct-bank raceway is PVC, usually Schedule 40, with Schedule 80 where the spec calls for the heavier wall or where the conduit comes out of the ground and needs the extra protection. PVC is cheap, it does not corrode in wet soil, and it solvent-welds into a continuous run fast. Verify the conduit type against the project spec, because some owners and some jurisdictions call for specific wall schedules, fiberglass, or concrete-encased steel in particular spots.
The layout is a grid, drawn on the duct-bank configuration detail: so many conduits wide and so many high, on a fixed center-to-center spacing, with each duct numbered to a circuit on the schedule. The drawing is not a suggestion. The position of every conduit in the grid drives the ampacity calculation, because where a conduit sits in the bank decides how much heat its neighbors dump on it.
Build spares into the grid from the start. A duct bank is the one raceway you truly cannot add to later, so the design carries empty conduits, often a quarter or more of the bank, capped and roped for the circuit nobody has drawn yet. Pulling a spare into an existing duct bank is a pull. Adding a duct to a poured bank is a demolition.
Spacers and the duct bank geometry
Manufactured duct-bank spacers hold the conduit grid in position while you build and pour. They come as a base spacer that sets the bottom row off the trench floor and intermediate spacers that stack the rows above it, snapping the conduits into a fixed center-to-center pattern. They are cheap, and they are the only thing standing between a clean grid and a pile of conduit that shifted when the concrete hit it.
Uniform spacing is not cosmetic. The ampacity study assumes the conduits sit at the spacing on the detail, commonly a minimum of around 3 in clear between conduits, so heat has room to move and the concrete can flow completely around each duct. Crowd two conduits together and you have created a hot spot the study never accounted for, and you have left a void where the concrete bridged instead of filling. Spread the spacers along the run at the spacing the manufacturer lists, often in the range of 5 to 8 ft, closer on the larger, heavier conduit that sags between supports.
Tie the spacers down. The spacer grid holds the geometry, but the whole assembly still wants to float, which is the next problem.
How deep does a duct bank go?
Minimum cover over a duct bank is set by NEC Table 300.5 for circuits up to 1000 V, measured from the top of the concrete to the finished grade, and it depends on what is above it. Concrete-encased conduit gets a shallower allowance than direct burial because the concrete is the protection, but under a road, a driveway, or anywhere subject to vehicular traffic the table commonly calls for 24 in of cover. Verify the exact figure against the adopted code edition, because the table has several columns and the number changes with the wiring method and the surface above.
Medium-voltage duct banks, above 1000 V, fall under the deeper cover requirements in the code and very often under a project spec that is deeper still, with 30 in or 36 in to the top being common on utility and data center work. The spec usually governs here, and it usually wins by going deeper than the code minimum.
Measure cover to the top of the concrete, not to the top of the conduit, and hold it across the whole run, including where the grade dips or a future road is planned. The shallow spot is the one a trencher finds two years later.
The concrete encasement
The encasement is the concrete jacket around the conduit grid, and the common spec is a minimum of 3 in of concrete on all sides of the outermost conduits. The NEC, for raceways approved for burial only where concrete-encased, calls for a concrete envelope of at least 2 in, but most engineered duct banks specify 3 in or more, so the project detail governs the thickness. The concrete is usually a lean mix, often specified around 3000 psi, and on a lot of jobs it is dyed red.
The red dye is a warning, not decoration. Red concrete underground tells the next crew with a backhoe that they have hit an electrical duct bank, not a random slab, before they hit a live feeder. It is cheap insurance against a dig-in.
Place the concrete so it fills every gap in the grid. The mix has to flow under and between the conduits, not bridge across the spacers and leave voids, so the slump and the placement matter and you work the concrete around the ducts as you go. On a tall bank you pour in lifts rather than dumping the full depth at once, both to control the float and to let the concrete consolidate around each row before the next load comes in.
Why do the conduits float in the pour?
Conduits float because PVC is far less dense than wet concrete, so the empty ducts want to rise like corks the moment the concrete goes around them. An unsecured grid lifts off the spacers, drifts out of position, and surfaces in the top of the pour, and now your carefully spaced bank is a tangle with the cover blown and the geometry gone. This is the single most common way a duct bank pour goes wrong.
You stop it by tying the bank down before the concrete comes. The common method is to drive rebar stakes into the trench floor on each side of the bank at intervals, then tie the conduit grid or the spacers to the stakes, often with a rebar bar across the top, so the assembly is anchored against the uplift. The spacers get tie-wired to each other and to the steel so the whole grid acts as one held-down unit.
Then you control the pour. Place in lifts, do not free-dump the full height, and if you vibrate, vibrate carefully, because over-vibration is exactly what frees the ducts to float. The tie-down and the staged pour work together. Skip either and you watch the bank come up out of the wet concrete with no way to push it back.
Rebar and the reinforced duct bank
Not every duct bank is reinforced, but the ones that span or carry a load are. When the bank crosses under a road and has to carry traffic, runs through fill that could settle, or spans an open cut where it acts like a beam, the engineer reinforces it with a rebar cage so the concrete does not crack and shear the conduits inside. At that point the duct bank is a structural element designed by the engineer, not just a protected raceway.
The reinforced bank is built like any other reinforced concrete: a cage of longitudinal bars and ties sized and spaced by the structural drawing, with the conduit grid held inside it and the specified cover maintained over both the steel and the ducts. The rebar that ties the conduits down for float control is not the same steel as the structural cage, so read the drawing for which one you are building.
Where the bank is reinforced, the inspection gets stricter, because now the steel placement, the cover over the rebar, and the concrete strength all matter to a structure, not just to a raceway. Build it to the structural detail, and do not improvise the cage to make the conduit grid fit.
Why is duct bank ampacity lower than a single conduit?
Conductors in a duct bank heat each other, so each one carries less current than the same conductor would alone in a single raceway. This is mutual heating, and it is the central electrical fact of duct bank design. Every loaded conduit is a heat source, the concrete and soil only shed that heat so fast, and a conduit buried in the middle of the grid is surrounded by other heat sources with no easy path out. The center duct in a large bank is the worst case and gets derated the hardest.
Because of this, you do not size duct-bank conductors off the basic NEC Table 310.16, which assumes a single raceway in a simpler condition. Duct-bank ampacity is engineered. For low-voltage work the NEC provides duct-bank figures in Annex B, and for medium voltage, 2001 V to 35000 V, the ampacity tables sit in 310.60, both built on the Neher-McGrath method that solved the heat-rise problem for buried cable systems. On a real data center bank the design usually comes from a full Neher-McGrath ampacity study, not a table lookup, because the configuration, the load factor, the burial depth, and the soil all move the number.
Verify the article and table numbers against the adopted code edition, and treat the engineered study and the project spec as the governing documents. Neher and McGrath showed that the center conductor in a tightly packed bank can derate toward 60 percent of its free ampacity, which is why the spacing on the detail is not negotiable.
- Mutual heating
- The temperature rise each loaded conductor causes in its neighbors, which lowers the ampacity of every conductor in the bank
- Neher-McGrath
- The 1957 calculation method behind engineered duct-bank ampacity and the NEC Annex B and 310.60 figures
- Annex B / 310.60
- The NEC duct-bank ampacity figures, Annex B for low voltage and 310.60 tables for 2001 V to 35000 V; verify against the adopted edition
Soil thermal resistivity (RHO) and thermal backfill
The ground around a duct bank carries the heat away, and how well it does that is its thermal resistivity, called RHO, in units of degrees C-cm per watt. Low RHO means the soil pulls heat out fast and the conductors run cooler, so they carry more current. High RHO means the heat stacks up and the ampacity drops. The ampacity study assumes a specific RHO, and that assumption is one of the biggest levers on the final conductor size.
A common default in the study is a soil RHO around 90, which is conservative for a lot of damp ground. Native soil that dries out, though, can climb much higher and choke the bank, which is why high-load feeders often sit in engineered thermal backfill instead of native spoil. A fluidized thermal backfill or a controlled thermal sand holds a low, stable RHO, sometimes near 50 to 60 even after it dries, and getting the heat out can lift ampacity by 10 to 30 percent versus poor native soil.
The catch is that the installed backfill has to match the study. If the design assumed a thermal backfill with a specified RHO and the field placed native dirt, the bank is undersized for its real thermal environment and nobody will know until the feeders run hot. Place the backfill the study called for, and confirm its RHO if the spec requires a thermal test.
- RHO
- Thermal resistivity of the soil or backfill in degrees C-cm per watt; lower RHO carries heat away and raises ampacity
- Thermal backfill
- Fluidized thermal backfill or thermal sand engineered to hold a low, stable RHO around the bank, replacing native spoil
Joints, solvent weld, and the slope to drain
PVC duct-bank conduit is joined by solvent welding: clean and prime the bell and the spigot, apply cement, and seat the joint so it fuses into one continuous, watertight run. Stagger the joints across the rows rather than landing them all at the same station, both for strength and so the bank does not have a plane of weakness where every coupling lines up. Use the long-radius bells and the manufactured sweeps the design calls for, not field-heated bends, because a kinked or flattened duct fails the mandrel later.
Water gets into a duct bank no matter what you do, so you slope it to drain. The bank is pitched along its length toward the manholes or handholes so any water that enters runs to a low point where it can be pumped or drained, instead of standing in the duct against the cable. A common target is a slope on the order of 3 in per 100 ft toward the access points, but verify the figure against the project spec, and never form a trap, a low belly between two high points, where water collects with nowhere to go.
The slope is set when you build the grid and it is locked when you pour. A duct bank poured dead level, or worse with a belly in it, holds water for its whole life.
Manholes, handholes, and the pull points
Manholes and handholes are the access points where the duct bank surfaces enough to pull, splice, and rack cable. They break a long run into pullable segments, and their spacing is a pulling decision as much as a layout one. Set them too far apart and the cable pull between them exceeds the tension or sidewall limits the cable can take. The spacing, the number of bends between pulls, and the resulting pulling tension are worked out together, which is the subject of the cable-pull plan that goes with this guide.
A manhole is sized for a person to get in and work, rack the cable on the walls, and make up splices. A handhole is the smaller version where a full manhole is not needed. Both get their own drainage, a sump or a drain in the floor, because they are the low points the duct bank slopes toward and water will collect there.
Inside, cable is racked on supports along the walls so it is held off the floor, supported through its bend, and reachable for testing and future work. The bend radius where the cable leaves the duct and turns into the rack is a real constraint, the same minimum bend radius the cable-pull guide covers, and it is easy to violate in the tight space of a manhole at the end of a hard pull.
Mandrel and proofing the ducts
Before any cable goes in, every duct in the bank gets proofed: you pull a mandrel through it to prove it is clean, round, and clear end to end. The mandrel is a rigid slug, commonly about 12 in long and sized roughly 1/4 to 1/2 in under the conduit inside diameter, so if it passes you know the duct holds its shape and has no crushed spot, no concrete intrusion at a joint, and no debris. A duct that fails the mandrel is rejected, and on a lot of specs a failure can reject the whole run, which tells you how seriously it is taken.
After the mandrel, the duct is usually swabbed with a brush or a rubber swab to clear grit, then a pull rope or pre-marked mule tape is left in for the cable that follows. Cap both ends. An open duct is an invitation for water, mud, and a curious animal, and a duct you proofed clean in the morning is full of trash by the time the cable crew arrives if it sat open.
Proof the duct before the cable, not the cable into an unproofed duct. A blockage found by a mandrel is a few minutes with a rod. The same blockage found by a stuck cable is a damaged pull in a duct you cannot open.
Stub-ups and the transition into the building
The duct bank has to come out of the ground and into the building, the equipment, or the manhole, and that transition is the stub-up. The conduits turn up out of the bank, usually on long-radius sweeps, and terminate where the cable needs to land, with bell ends on the duct mouths so the cable is not dragged over a sharp conduit edge as it feeds out. The bend radius on the stub-up sweep has to respect the cable's minimum bend radius, the same limit the cable-pull guide covers, because a tight stub-up crushes the cable right where it leaves the protection of the bank.
Spare ducts get stubbed up and capped along with the working ducts, so the spare is reachable, not buried blind under a slab. A spare you cannot find is not a spare.
The stub-up is also where the duct bank meets a different wall or floor, so the penetration gets sealed against water and, where required, firestopped or made gas-tight. Water travels down a sloped duct bank, and a stub-up into a building is a place it can come up into the room if the duct is not sealed, which is how a duct bank floods an electrical room nobody thought to waterproof.
Tracer wire and warning tape
A concrete-encased PVC duct bank is invisible to a metal locator, because there is no metal in it to find, so you give the locator something to read. A tracer wire, run the length of the bank and brought up at the access points, lets a crew locate the nonmetallic duct bank from the surface later. Without it, the only record of where the bank runs is the as-built drawing, and the only way to find it in the field is to dig.
Above the bank, in the trench backfill, you lay a warning tape, usually red and printed to say there is a buried electric line below. Many specs call for detectable tape with a metallic core so it doubles as a locator target. The tape sits well above the bank so a digger hits the tape, and stops, before the shovel reaches the concrete.
Mark it and record it. The tracer wire and the tape protect the bank in the field, and the as-built drawing records where it actually went, which is rarely exactly where the plan drew it. Capturing the real location, the real depth, and the duct assignments is its own discipline, and a duct bank with no usable as-built is a duct bank the next crew finds with a backhoe.
Pour-day execution
Pour day is the point of no return, so the last hour before the concrete is the most valuable inspection on the whole job. Walk the bank one final time and confirm the things you cannot fix afterward: the spacing is right and uniform, the tie-downs are in and tight so nothing floats, every duct has been mandrel-proofed and capped, the slope is true with no belly, the cover will be made over the whole run, and any rebar cage is placed and tied. This is the last look. Once the truck backs up, the duct bank is whatever it is right now, forever.
Place the concrete in lifts up the height of the bank rather than dumping the full depth, working the mix around and under each conduit so it fills the grid without voids and without freeing the ducts to float. Keep an eye on the bank as it fills, because a tie that lets go or a spacer that shifts shows itself in the wet concrete while you can still react.
Pour in weather the mix can handle, protect it as you would any structural concrete, and do not bury a problem you noticed and hoped would be fine. The concrete does not get more honest after it sets.
What does the inspector check before the pour?
The pre-pour inspection on a duct bank is often a formal hold point, a special inspection or an engineer's sign-off the job cannot pour past until it is signed. The inspector is checking the things the concrete is about to hide forever: the conduit spacing against the detail, the tie-down against float, the cover that will be achieved, the slope to the manholes, the mandrel-proof records, and on a reinforced bank the rebar placement and cover. Treat it as the gate it is and have the bank truly ready, because a failed pre-pour inspection is cheap and a poured-over defect is not.
Acceptance does not end at the pour. After the cable is pulled, medium-voltage cable gets tested, commonly an insulation-resistance megger and often a higher-voltage acceptance test such as a VLF or DC withstand per the project and the testing standard, to prove the cable came through the pull and the install undamaged. That testing is its own subject, but it is the proof that closes out the run.
The clean way through both gates is records. The mandrel proofs, the spacing and cover checks, the pour documentation, and the cable test results are what turn an inspection into a signature instead of an argument.
Common failures and how they show up
The duct bank failures are a short list, and almost all of them trace to something skipped before the pour. The bank that floated in the concrete is the classic: ducts not tied down, conduits surfaced in the pour, spacing gone and cover blown, and no fix short of breaking it out. A duct never mandrel-proofed shows up as a blockage found at the worst possible time, when a cable stalls halfway through a pull in a duct that was crushed at a joint or had concrete intrude.
Spacing built wrong is a quiet failure. The bank looks fine, but the conduits sit closer than the study assumed, the mutual heating is worse than designed, and the feeders run hotter than their ampacity allows for the life of the bank. Short cover is the same kind of hidden problem until a trencher or a road load finds the shallow spot.
No slope, or a belly in the run, leaves ducts standing full of water against the cable. And no tracer wire or warning tape leaves the next crew with a nonmetallic bank they cannot find until they dig into it. None of these are equipment problems. They are install problems, and every one of them was preventable before the concrete set.
What to document
A duct bank you cannot reconstruct from paper is a duct bank the next crew has to discover with a shovel. The record proves the bank was built to the design and gives the people who come after it the location, the depth, the duct assignments, and the proof that each duct is clear. It is also what closes out the inspection and the cable testing.
For each run, capture the conduit type, size, and count, the grid spacing, the cover achieved, the encasement thickness and concrete used, the slope, and the mandrel-proof result per duct, along with the as-built horizontal and vertical location. Tie the duct numbers to the circuits and note the spares. The as-built is the document that outlives everyone on the job.
| Field to record | Why it matters |
|---|---|
| Run and conduit type, size, count | Identifies the bank and ties ducts to circuits |
| Grid spacing, center to center | Drives the mutual-heating ampacity |
| Cover over the bank | Proves the depth meets the table and the spec |
| Encasement thickness and concrete | Confirms the jacket and the strength placed |
| Slope and drainage direction | Shows water drains to the manholes, no trap |
| Mandrel proof per duct | Proves each duct is clear and round |
| As-built location and depth | Lets the next crew find the bank without digging |
Common mistakes
- Leaving the conduit grid untied, so the ducts float up out of the concrete during the pour.
- Pouring without mandrel-proofing every duct, then finding a blockage when a cable stalls in the pull.
- Building the spacing tighter than the ampacity study assumed, so the feeders run hotter than designed.
- Coming up short on cover, especially under a road or at a grade dip nobody checked.
- Pouring the bank level or with a belly, so the ducts hold water against the cable for life.
- Sizing duct-bank conductors off Table 310.16 instead of the engineered duct-bank ampacity.
- Placing native spoil where the study called for thermal backfill, undersizing the bank's heat path.
- Skipping the tracer wire and warning tape on a nonmetallic bank nobody can locate later.
- Dumping the full pour height at once and over-vibrating, which frees the ducts to float.
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 framework lives in the NEC, NFPA 70. Cover over an underground duct bank comes from Table 300.5 for circuits up to 1000 V, with the deeper requirements for medium voltage above 1000 V, both measured to the top and both adopted and amended by jurisdiction. Duct-bank ampacity, the part that makes a duct bank different from a single raceway, comes from the engineered figures: Annex B for low voltage and the 310.60 tables for medium voltage, 2001 V to 35000 V, both grounded in the Neher-McGrath method. Conduit fill for the conductors inside the ducts is Chapter 9, the subject of a separate guide. Confirm every article and table number against the adopted code edition before you cite it on a submittal, because the numbers move between cycles.
The ampacity engineering itself traces to the Neher-McGrath paper and the IEEE and IEC cable-ampacity methods the study software is built on, and on a real job the governing document is the project's engineered duct-bank design and ampacity study, which fixes the configuration, the RHO, the backfill, and any reinforcement. The PVC conduit is made to the NEMA TC standards for rigid nonmetallic conduit, and the concrete to the project's structural specification.
Where the engineered design or the spec is stricter than a code minimum, the design and the spec govern. The cable testing that closes out the run follows the project and the acceptance-testing standard, and the medium-voltage cable's own minimum bend radius and pulling limits come from the manufacturer, which the cable-pull guide covers.
Units, terms, and conversions
A duct bank carries terms from concrete, from the cable ampacity world, and from the code, and they show up across the structural detail, the electrical drawing, and the ampacity study. Cover and spacing are in inches, slope in inches per 100 ft, concrete strength in psi, and thermal resistivity in degrees C-cm per watt, which a metric study may give in K-m per watt instead.
The terms below are the ones that travel between the duct-bank detail, the ampacity study, and the inspection.
- Duct bank
- A grid of conduits run underground and encased on all sides in concrete to protect and route multiple feeders
- Encasement
- The concrete jacket around the conduit grid, commonly 3 in on all sides, often dyed red for identification
- Spacer
- Manufactured base and intermediate pieces that hold the conduit grid at the specified center-to-center spacing
- Mutual heating
- The rise in temperature each loaded conductor causes in its neighbors, which lowers duct-bank ampacity
- RHO
- Thermal resistivity of the soil or backfill in degrees C-cm per watt; lower RHO carries heat away and raises ampacity
- Neher-McGrath
- The calculation method behind engineered duct-bank ampacity and the NEC Annex B and 310.60 figures
- Mandrel
- A rigid slug pulled through each duct to prove it is clean, round, and clear before cable
- Tracer wire
- A conductor run with a nonmetallic duct bank so it can be located from the surface later
FAQ
How deep does a duct bank go?
Minimum cover comes from NEC Table 300.5 for circuits up to 1000 V, measured to the top of the concrete. Under a road or vehicular traffic the table commonly calls for 24 in. Medium-voltage banks and project specs often go deeper, 30 in to 36 in. Verify against the adopted code edition and the spec.
Why is duct bank ampacity lower than a single conduit?
Conductors in a duct bank heat each other, so each carries less current than it would alone, and the center duct is derated hardest. You size from engineered figures, NEC Annex B for low voltage and the 310.60 tables for medium voltage, built on Neher-McGrath, not from Table 310.16. A full ampacity study governs real banks.
Why is duct bank concrete red?
The concrete is dyed red as a warning. Red concrete underground tells the next crew with a backhoe that they have hit an electrical duct bank, before they hit a live feeder. It is a cheap way to prevent a dig-in. Many specs also require a warning tape and a tracer wire above the bank.
Why do conduits float in the concrete pour?
PVC is far less dense than wet concrete, so empty ducts rise like corks when the concrete goes around them. An untied grid lifts off the spacers and surfaces in the pour, blowing the spacing and the cover. Tie the grid down to staked rebar, pour in lifts, and do not over-vibrate, which frees the ducts to float.
Do I need thermal backfill around a duct bank?
You need it when the ampacity study assumed it. High-load feeders often sit in fluidized thermal backfill or thermal sand, which holds a low, stable RHO near 50 to 60 even when dry, instead of native soil that can dry out and choke the bank. Place what the study called for, or the feeders run undersized.
What happens if a duct fails the mandrel test?
A duct that will not pass the mandrel has a crushed spot, a concrete intrusion at a joint, or debris, and it is rejected for cable. On many specs a single failed duct can reject the whole run. Proof every duct before cable, swab it, leave mule tape, and cap both ends so it stays clean.
What conduit is used for a concrete-encased duct bank?
PVC, usually Schedule 40, is the common duct-bank raceway, with Schedule 80 where the spec wants the heavier wall or where the conduit exits the ground. PVC does not corrode in wet soil and solvent-welds into a continuous run. Verify the type against the project spec, since some owners require fiberglass or specific schedules.
How much spacing goes between duct bank conduits?
The configuration detail sets it, commonly a minimum around 3 in clear between conduits, held uniform across the grid. The spacing is what the ampacity study assumed, so crowding conduits closer makes the mutual heating worse than designed. Manufactured base and intermediate spacers hold the grid, placed roughly every 5 to 8 ft along the run.
Why does a duct bank need to slope to the manholes?
Water gets into a duct bank no matter what, so the bank is pitched to drain to the manholes or handholes, where it can be pumped out instead of standing against the cable. A common target is about 3 in per 100 ft, but the spec governs. Never form a belly or trap that holds water.
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.