Datacenter
Busway receiving and megger QA for data center power
Receive busway dry, torque every joint to the indicator, and megger phase-to-phase and phase-to-ground before you energize the run to the racks.
Direct answer
Busway is a prefabricated metal-enclosed run of bus bars that distributes high-ampacity power overhead to data center PDUs, RPPs, and racks. Receiving and QA live at the bolted joints and the insulation-resistance, or megger, reading. Test phase-to-phase and phase-to-ground, record a baseline, and verify joint torque. Manufacturer instructions and NETA acceptance testing govern.
Key takeaways
- Busway QA lives at the bolted joints and the megger reading: test phase-to-phase and phase-to-ground, record a baseline, and verify joint torque.
- A low busway megger reading is moisture far more often than damage; dry the run and re-read before condemning it.
- Modern single-bolt busway joints torque to the manufacturer value (commonly around 70 lb-ft / 95 N-m) until the twist-off head shears or indicator shows.
- For busway rated 600 V and below a 1000 Vdc megger test is common; medium-voltage bus tests at roughly 2500 to 5000 Vdc.
- ANSI/NETA ATS governs acceptance; the IR minimum scales to actual run length and the manufacturer's instructions override any general value.
What busway is, and why the joints and the megger are the QA
Busway, also called bus duct, is a prefabricated run of copper or aluminum bus bars in a metal housing, built in straight lengths and fittings that bolt together into a continuous power path. It replaces large parallel cable runs with a rigid, high-ampacity conductor you hang from the structure. In a data center it carries power overhead from a switchboard, a PDU, or a remote power panel down the rack rows, and the rack power taps off it through plug-in units.
Two things decide whether that run is any good, and neither is the bar itself. The first is the bolted joint, where one length meets the next, because a loose joint runs hot and heat is how busway fails. The second is the megger reading, the insulation resistance between the bars and to ground, because that number tells you whether the insulation is dry and intact before you put voltage on it over the racks.
The bar is rarely the problem. The joints and the moisture are. That is why the inspection weights the joint torque and the megger over everything else, and why both have to be recorded as a baseline you can read the next test against.
Why busway feeds the racks instead of cable
Data centers run busway overhead because the load moves and the cable plant cannot. A white-space row gets re-racked, densified, and re-fed on a cycle measured in months, and a continuous overhead bus lets you add or relocate a rack tap without pulling new feeder cable back to the panel. You plug a new tap into the bus where you need it and you are done.
Ampacity is the other reason. A single busway run carries hundreds to a few thousand amps to a row of rack PDUs or remote power panels, and the same capacity in cable means many parallel sets and a congested tray. The bus is compact, it is rigid, and the rating is stamped on the nameplate rather than calculated across paralleled conductors.
The tradeoff is that all that capacity rides through bolted joints every few feet, and the whole run shares one insulation system. Get a joint or the insulation wrong and you have put a single high-current fault point directly above the racks. The flexibility that makes busway worth it is exactly what concentrates the risk at the joints, so the QA goes where the risk is.
Feeder busway, plug-in busway, and the plug-in units
There is no single busway, so read the type off the submittal before you write the inspection. Feeder busway is the high-ampacity trunk with no tap openings. It moves power from the source to a distribution point with the lowest drop and the fewest joints, and it is what you run where you do not need to tap along the way.
Plug-in busway has tap openings along its length at a fixed spacing, usually every couple of feet, with covers over the unused ones. It is what runs down the rack rows so you can land power where the racks actually are. The plug-in unit is the tap: a bolt-on or stab-in device, often a fused switch or a molded-case breaker in a small housing, that clamps onto the bus through an opening and feeds a branch out to a rack PDU or a remote power panel. The unit is where the interlock and the stab engagement live, and it is a separate inspection from the run.
Most white-space busway is indoor, dry-location rated. Outdoor or wet-location bus is a different listing with a different gasket and drain story, so confirm the location rating against where the run actually goes. A length destined for a dry hall and a length destined for a roof feeder are not interchangeable, and the substitution shows up here.
Receiving busway at the dock
Inspect busway the moment it lands, before you sign the bill of lading, because the cheapest place to catch shipping damage is on the truck and the claim window starts closing at delivery. Busway ships as straight lengths, elbows, tees, tap boxes, and the plug-in units, often across several pallets, and the joint hardware and end closures ship loose. Count all of it against the packing list while the driver is there.
Walk the housing for dents, crushed ends, racked or twisted sections, and paint scraped to bare metal, because a housing that took a hit can pinch the bars or crack the insulation inside where you cannot see it. Check the joint ends hard. A bent end plate or a damaged joint stack means that length will not make up a good joint no matter how you torque it, and that is a defect you want on the carrier, not on your schedule.
Note any damage as a specific exception on the bill of lading, by piece mark and location, and photograph it before you unload. Dented housing on length B-3, two-inch gouge at the joint end, beats a vague note that loses the claim. Sign clean over visible damage and the carrier's first move is to say it happened on your dock under your rigging. Visible damage goes on the BOL now. The formal freight claim has a longer deadline, but the notation at delivery is what keeps it alive.
Why you store busway dry and keep the ends capped
Busway insulation is hygroscopic. The materials between and around the bars pull moisture out of the air, and moisture in the insulation is the classic busway problem, because it drives the megger reading down and, left in service, it becomes a flashover path. The whole storage discipline exists to keep water out of a run you will not energize for months.
Store it indoors, dry, and off the ground, oriented as the manufacturer directs, with the ends kept capped. The shipping end caps are there to keep dust and moisture out of the bus stack, so leave them on until you actually make the joint. A length stored uncapped in a damp warehouse, or worse outdoors under a tarp that sweats, will read low when you finally megger it, and then you cannot tell transit damage from storage neglect.
Keep the loose joint kits, end closures, and plug-in units dry and labeled too. The expensive version of this mistake is a run stored wet all winter, meggered low in spring, and condemned as defective when all it needed was drying out. Log the storage conditions, because a dry-storage record and a baseline megger are what back you up when the question is whether the gear was preserved.
Why is my busway megger reading low?
A low busway megger reading is moisture far more often than it is damage. The insulation is hygroscopic, so a run that sat in a humid space, got rained on, or came in cold and sweated when it hit a warm room reads low because there is water in and on the insulation, not because the insulation is bad. Wet busway is the number one reason a good run gets wrongly condemned at acceptance.
Before you blame the bus, rule out the easy causes. A reading taken on cold gear, below the dew point, or with the surfaces dirty or damp will be low and meaningless. Surface tracking across a dirty joint or a wet end reads as a fault that wiping and drying makes disappear. Temperature alone swings the number hard, so a cold run reads lower than the same run warm, and that is why you correct to a standard temperature before you compare anything.
If the run is genuinely wet, dry it out and read it again before you make any call. A reading that climbs as the bus dries is the proof it was moisture, not a defect. A reading that stays low after a real dry-out is when you start looking for a damaged insulator, a pinched bar, or a cracked support. Condemning a wet run is the rookie call. Drying it first is the experienced one.
Drying out busway before you trust the reading
When a run reads low and you have confirmed it is moisture, dry it before you megger it for record and before you energize. Heat drives the water out of the insulation and the reading recovers. The methods run from simple to involved: move the run into a warm dry space and let it stabilize, apply external heat with blowers or heat lamps kept at a safe distance, follow the manufacturer's published drying procedure, or in the worst case send a controlled low current through the bars to warm them from the inside. The manufacturer's instructions govern the method and the temperature limit, because too much heat damages the insulation you are trying to save.
Watch the megger while it dries. A useful field tell is that the resistance often dips before it climbs, because heat first mobilizes the moisture and drives it toward the surface before it leaves the insulation. Take readings on a schedule and plot them. When the number climbs, levels off, and holds, the insulation is dry. Use a lower test voltage in the early, wet stage so you do not stress insulation that is still damp.
Do not energize wet busway to dry it out by accident. Putting full voltage across damp insulation is how you turn a drying problem into a flashover. Dry it the controlled way, prove it with a rising reading and a good polarization index, then read the baseline.
Installation QA: supports, alignment, and the run layout
Busway is a rigid conductor that has to be supported as a structure, not draped like cable, so the install QA starts with the hangers. Support the run at the spacing in the manufacturer's instructions, commonly on the order of every five to ten feet for horizontal runs and at every floor for risers, and never let a joint carry weight or hang in space between two supports. A joint should land near a hanger, not bridge a gap.
Check alignment before you torque anything. Lengths have to come together square and in line, because a joint pulled into place against a misaligned run is a joint under mechanical stress for its whole life, and stress plus heat is how it loosens. Set the supports so the bus is straight and level, the sections meet without forcing, and the run carries its expansion fittings where the layout and the building movement call for them. A long horizontal run and every riser has to absorb thermal growth, and the expansion fitting is what gives it room.
Verify phasing as you go. The bars sit in a fixed phase sequence, and a length or a fitting installed flipped puts the phases out of rotation at that joint. Confirm phase orientation at each joint and end to end before energizing, not after a motor on a downstream load runs backward.
What torque do busway joints take?
Busway joint torque comes from the manufacturer, and on most modern busway the joint is a single bolt with a built-in torque indicator, not a row of bolts you torque one by one. The single-bolt joint runs one high-strength bolt through a stack of conical spring washers that spread a few thousand pounds of clamping force across the bar contact. The indicator tells you when it is right: many designs use a twist-off head that shears at the correct torque, or a visual marker that changes as the bolt reaches tension. When the head snaps or the indicator shows, that joint is made.
Do not guess and do not borrow the value from another product. A common single-bolt value lands around 70 lb-ft, but the number, the bolt, and the indicator all belong to the specific busway, so torque to the manufacturer's published value with a calibrated wrench when there is no twist-off head. Where the joint is a bolted-bar type with multiple bolts, torque each to the table and mark it.
Here is why this is the heart of the inspection. A loose joint is the failure point. It runs hot under load, the heat relaxes the joint further, and the cycle ends in a burned joint and an outage over the racks. After torque, mark each joint so a later thermographic scan and the next inspector can see it was verified. A joint you cannot prove was torqued is a joint you have to assume was not.
How do you megger a busway run?
Megger a busway run with a megohmmeter, phase-to-phase between every pair of bars and phase-to-ground from each bar to the housing, with the run de-energized, isolated, and grounded before and after. Take the reading at one minute and record it. The housing is your ground reference and the equipment grounding path, so phase-to-ground is read to the enclosure.
Match the test voltage to the system, from the manufacturer's published data or the NETA table of test values. For busway rated 600 V and below, a 1000 Vdc test is common; medium-voltage bus is tested higher, often in the 2500 to 5000 Vdc range per the table. On a long run, take a polarization index, the ratio of the ten-minute reading to the one-minute reading, where a rising ratio says the insulation is dry and a flat or falling ratio points to moisture or contamination. Record temperature and relative humidity with every reading and correct to a standard temperature, commonly 20 degrees C, because insulation resistance swings hard with both and an uncorrected number is not comparable to the next one.
Read it as a baseline, length by length and across the made joints, and keep it. NETA acceptance testing gives a minimum scaled to the run length, but the trend matters as much as the single number. The baseline at install is what the next megger is read against, and a drop over time tells you something changed in the insulation before it tells you with smoke.
- Insulation resistance (IR)
- The megohmmeter reading across insulation, phase-to-phase and phase-to-ground, recorded with temperature and humidity
- Polarization index (PI)
- Ratio of the ten-minute to the one-minute IR reading; a rising ratio indicates dry, sound insulation
- Megohmmeter (megger)
- The instrument that applies a DC test voltage and reads insulation resistance in megohms or gigohms
What insulation resistance is acceptable for busway?
Acceptable busway insulation resistance is whatever the manufacturer's published minimum and the NETA acceptance table call for, read against a dry baseline, not a single magic number you carry in your head. NETA acceptance testing scales the minimum to the run length, because a longer run has more insulation surface in parallel and reads lower for the same condition. Recent NETA acceptance editions moved to using the actual measured length in the formula rather than a fixed nominal length, so use the run you actually have.
In practice a clean, dry low-voltage run reads in the high megohms to the gigohms, and a reading in the low megohms or below on a dry run is a flag, not a pass. Do not accept a number near the floor without finding out why. Check temperature, humidity, and surface cleanliness, dry the run if it is wet, and re-read.
The pass is the corrected reading against the manufacturer and NETA minimum, with a polarization index that says the insulation is dry, recorded as the baseline for the life of the run. A run that just clears the floor is a run that has no margin, and on busway over racks the margin is the point.
Continuity, phasing, and ground bonding along the run
Before energizing, prove three things electrically beyond the insulation reading: that each phase is continuous end to end, that the phases are in the right rotation, and that the housing is bonded as a continuous equipment ground. Continuity confirms every joint actually made up and no bar is open. Run a continuity or low-resistance check phase by phase from one end of the completed run to the other.
Phasing has to be right end to end and consistent across the taps. Verify phase rotation at the source, along the run, and at each plug-in location, because a tap pulled off the wrong orientation feeds a rack PDU out of phase, and that surfaces as a downstream problem you will chase from the wrong end. Phase rotation matters wherever three-phase loads hang off the bus.
The housing is the equipment grounding conductor on most busway, so its continuity is a safety item, not a formality. Confirm the ground bus or the housing-as-ground is electrically continuous across every joint, because that is the fault-current path that lets the breaker clear a ground fault. A joint that is mechanically together but has a poor ground bond is a path that will not be there when a fault needs it. Bond the run to the building ground at the points the design calls out and verify the bond, not just the presence of a wire.
Contact resistance across joints with a ductor
Where the spec calls for it, measure the contact resistance across each bolted joint with a low-resistance ohmmeter, the ductor, which pushes a known DC current through the joint and reads the millivolt drop to give a micro-ohm value. The joint resistance tells you whether the bolted connection is actually making good metal-to-metal contact, which torque alone does not fully prove, because a properly torqued joint over a dirty or oxidized contact surface still reads high.
Read each joint and compare. The numbers across like joints on the same run should sit close to each other and close to the manufacturer's expected value where one is published. A joint that reads markedly higher than its neighbors is the one that will run hot, and it gets pulled, cleaned, and remade before energizing, not flagged for later. The ductor is the test that catches the joint the torque wrench and the eye both passed.
This is an acceptance-level check, not a step on every job, so follow the project specification and the NETA acceptance scope for whether contact-resistance testing on the bus joints is required. Where it is required, it pairs with the torque verification: torque proves clamping force, the ductor proves the contact those clamped surfaces actually make.
Energizing the run and the infrared scan under load
Energize the busway only after the insulation reading, the continuity, the phasing, the joint torque, and the ground bond are all signed off, and bring it up in a controlled sequence: source first, then the bus, then the taps and the loads, watching the gear as you go. A run that passed every static test still gets watched the first time it carries real current.
The check that earns its keep here is the infrared scan under load. Once the run is energized and loaded to a meaningful fraction of its rating, scan every joint and tap with a thermal imager, because the joints are the hot spots and a loose or high-resistance joint shows up as a temperature rise long before it fails. A joint running hotter than the bar around it and hotter than its sister joints is the one to schedule for shutdown and remake. Scan through the inspection windows or covers the manufacturer provides, and record the load at which you scanned, because a temperature rise only means something against the current that caused it.
This is where the torque marking pays off. The scan finds the hot joint and the mark tells you whether it was verified at install, which points you at a workmanship miss versus a connection that loosened in service. Repeat the scan as part of the maintenance cycle, not just at acceptance, because the joint that reads clean cold at 60 percent load can be the one that drifts hot at full load a year in.
Plug-in unit QA: interlocks, stabs, and the branch to the rack
The plug-in unit is its own inspection, because the tap is where the run becomes rack power and where the moving parts live. Check that the unit's stabs or jaws engage the bus fully and squarely when it is installed, because a partial stab engagement is a high-resistance, high-heat connection in the one spot you cannot see once the cover is on. Then megger the unit and its branch the same way you megger the run, phase-to-phase and phase-to-ground, before it feeds the rack PDU.
Many plug-in units carry an interlock that prevents installing or removing the unit under load and keeps the door from opening on a closed switch. Confirm the interlock works and is not defeated, because that is the operator-safety feature on a device people will pull and replace in a live row over the life of the building. Verify the unit's overcurrent rating matches the branch and the downstream PDU, the phasing into the PDU is right, and the unit is fully seated and latched.
The branch from a plug-in unit to a rack PDU is a short run, but it is the last connection before the load, so it gets the same continuity and phasing proof as the trunk. Document each plug-in location by its position on the run, because a tap is identified by where it lands on the bus, the same way a joint is identified by its sequence.
The receiving and test packet, tied to run, joint, and tap
Every finding in the packet ties to a coordinate on the run, because a finding without a location is a finding nobody can act on. Busway is already a numbered structure: number the runs, the lengths and fittings by piece mark, the joints in sequence, and the plug-in locations by position, and key every photo, megger reading, torque verification, and exception to that coordinate. A note that reads Run B, Joint 7, infrared 41 degrees C rise at 60 percent load, photo 18, walks the next person straight to the joint. A note that reads one joint runs hot does not.
The packet pulls the whole job into one defensible record: the receiving exceptions on the bill of lading, the storage and dry-out log, the joint torque verification, the baseline megger with its conditions, the continuity and phasing proof, the ground-bond check, any ductor readings, and the energized infrared scan with its load. The table below is the spine that ties each item to a run coordinate, an evidence photo, a responsible party, and a status, the same way a switchgear receiving packet keys every finding to a lineup section.
| Run / joint / tap | Check | Evidence | Status |
|---|---|---|---|
| Run B, all lengths | Baseline megger, ph-ph and ph-gnd | IR sheet, 21 C / 45% RH | Recorded, in band |
| Run B, Joint 7 | Single-bolt torque, indicator sheared | Photo 18 | Verified |
| Run B, Joint 12 | Infrared 41 C rise at 60% load | Thermal image 06 | Open, remake scheduled |
| Run B, end feed | Ground-bond continuity to housing | Low-resistance reading | Verified |
| Tap B-4 | Plug-in stab engagement, interlock | Photo 23 | Verified |
| Length B-3 | Dented housing, noted on BOL | Photo 02, BOL exception | Carrier claim filed |
Field example: a plug-in run down a data hall row
Take a 1200 A plug-in busway run feeding one row of racks in a data hall, fed from a remote power panel at one end, eight straight lengths, seven joints, and six plug-in units out to the rack PDUs. The run arrived over the winter and sat in a conditioned but humid storeroom for three months with the ends capped.
At install the crew supports the run on the manufacturer's hanger spacing, brings each length together square, and makes each joint with the single-bolt connector until the twist-off head shears, marking every joint after. The first megger reads low, single-digit megohms phase-to-ground, which on a dry run that length would be a fail. Rather than condemn it, the tech checks conditions, finds the bus cold and the storeroom damp, warms and dries the run, and re-meggers. The reading climbs into the gigohms and a ten-minute polarization index confirms it was moisture, not damage. That corrected reading goes in as the baseline.
After continuity, phasing end to end, and a ground-bond check across every joint, the run is energized and loaded. The infrared scan finds Joint 5 running about 40 degrees C above the adjacent bar at 60 percent load. It is shut down, opened, and the contact surface is found lightly oxidized from the damp storage. Cleaned and remade, it scans flat. The packet closes with the dry-out logged, every joint torque-verified and marked, the baseline megger recorded, and Joint 5 documented as found and corrected. The win is plain: wet busway was dried instead of condemned, and the one bad joint was caught on the imager instead of in a burndown.
| Run / joint / tap | Check | Result | Status |
|---|---|---|---|
| Run, all lengths | Initial megger, ph-gnd | Low, single-digit megohms | Investigated, not condemned |
| Run, all lengths | Re-megger after dry-out | Gigohms, PI rising | Recorded as baseline |
| Joints 1-7 | Single-bolt torque | Twist-off heads sheared, marked | Verified |
| Joint 5 | Infrared at 60% load | 40 C rise vs adjacent bar | Remade, scans flat |
| Taps 1-6 | Stab engagement, interlock | Full engagement, interlock works | Verified |
| Run | Phasing end to end | Correct rotation | Verified |
What to document
Build the packet so that a question a year out gets answered from the file and not from somebody's memory of the run. Capture enough that someone who was never on the run can reconstruct what arrived, in what condition, how it was stored and dried, how each joint was made, and what the insulation and the energized scan read, all against the run coordinate the rest of the record keys to.
Record the run and piece-mark numbering, the bill of lading with its exceptions, the storage and dry-out log with conditions, the joint torque verification by joint, the baseline insulation-resistance readings with temperature and humidity and the polarization index, the continuity and phasing proof, the ground-bond check across the joints, any ductor contact-resistance readings, and the energized infrared scan with the load it was taken at. Where a joint was found hot and remade, record it as found and corrected. The table below is the minimum spine.
| Field to record | Why it matters |
|---|---|
| Run and piece-mark numbering | Every other record keys to it; without it the chain breaks |
| BOL and receiving exceptions | The legal record that preserves the freight claim |
| Storage and dry-out log | Proves the hygroscopic insulation was kept and brought dry |
| Joint torque verification | Documents the connection that fails when it is loose |
| Baseline megger with conditions and PI | The reference every later IR reading is judged against |
| Continuity and phasing proof | Confirms every joint made up and rotation is right |
| Ground-bond continuity across joints | Proves the fault-current path the breaker depends on |
| Ductor contact resistance, where required | Catches the high-resistance joint torque alone passes |
| Energized infrared scan and load | Finds the hot joint under real current, against the load |
Common mistakes
- Meggering wet or cold busway, reading it low, and condemning a good run that only needed drying out.
- Skipping the baseline megger at install, so there is no dry reference to read the next test against.
- Storing busway uncapped, outdoors, or under a sweating tarp, then discovering wet insulation at acceptance.
- Treating the single-bolt joint as torqued without verifying the twist-off head sheared or the indicator showed, and leaving joints unmarked.
- Borrowing a torque value from another busway product instead of the manufacturer's published value for the bus on the job.
- Letting a joint bridge a gap between supports or pulling a misaligned run together, so the joint lives under mechanical stress.
- Energizing before continuity, phasing, and ground-bond continuity across every joint are proven.
- Skipping the energized infrared scan, so a loose or high-resistance joint runs hot undetected until it burns.
- Installing a plug-in unit with partial stab engagement or a defeated interlock, hiding a high-heat connection under the cover.
- Filing photos and readings with no run coordinate, so a later finding cannot be tied to a joint, a tap, or a fix.
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 acceptance testing framework is ANSI/NETA ATS, the Standard for Acceptance Testing Specifications for Electrical Power Equipment, which carries a section for metal-enclosed busways covering the visual and mechanical inspection, the bolted-connection and torque verification, the insulation-resistance test phase-to-phase and phase-to-ground, and contact-resistance testing where applicable. Recent ATS editions revised the busway insulation-resistance acceptance to use the actual run length in feet rather than a fixed nominal length, so use the edition the project specifies and the run you actually have. The insulation-resistance test voltages and minimum values come from the manufacturer's published data or the NETA table of test values.
The product standards depend on the type. Feeder and plug-in busway rated 600 V and below is built and listed to UL 857, the busway standard, with the design, testing, and application guidance in the NEMA BU 1 series and the handling, installation, and maintenance guidance in NEMA BU 1.1. Installation and routine maintenance of those busways, including procedures after a fault or water immersion, are covered by NECA 408, the Standard for Installing and Maintaining Busways. Metal-enclosed bus in the higher-voltage, non-segregated, segregated, and isolated-phase forms is covered by IEEE C37.23, which is a separate family from the UL 857 busways and applies to the metal-enclosed bus that connects switchgear rather than the plug-in bus over the racks.
Electrical equipment maintenance and the storage and preservation expectations are framed by NFPA 70B, the Standard for Electrical Equipment Maintenance, with the understanding that the manufacturer's instructions govern over any general reference for torque values, test voltages, drying procedures, and storage. Confirm the applicable editions, the rated values, and the test parameters against the approved submittal and the manufacturer's documents, because the specific requirement is set by the project and the listing, not by the rule of thumb.
Units, terms, and conversions
Busway QA crosses ratings, test units, and a few trade synonyms, so the same item reads differently across a submittal, a nameplate, and a test report.
Busway and bus duct are the same thing, and the bars inside are the bus. Current ratings are in amps for the continuous rating and kA or kAIC for the short-circuit and interrupting ratings, so a run has both a continuous ampacity and a withstand rating, and both come off the nameplate against the submittal. Voltage is in V for low-voltage bus and kV for medium-voltage bus, given as the rated class, not the operating voltage. Insulation resistance reads in megohms or gigohms, always recorded with the temperature in degrees C or F and the relative humidity, because the reading moves hard with both. Torque is given in lb-ft or in-lb in US units and N-m in metric, and a single-bolt joint value of around 70 lb-ft is roughly 95 N-m, but the manufacturer's number governs. The run, length, joint, and plug-in position are not units, but they are the coordinate the whole packet turns on, so set the numbering once and use it everywhere.
- Busway / bus duct
- A prefabricated metal-enclosed run of bus bars in straight lengths and fittings, bolted into a continuous power path
- Feeder vs plug-in busway
- Feeder has no tap openings and moves power point to point; plug-in has tap openings along its length for rack power
- Plug-in unit
- A bolt-on or stab-in tap, often a fused switch or breaker, that clamps onto plug-in busway to feed a branch
- Single-bolt joint
- A busway joint clamped by one high-strength bolt and conical spring washers, often with a twist-off or visual torque indicator
- kAIC
- Thousand amps interrupting capacity, the short-circuit rating of low-voltage gear and overcurrent devices
- Ductor
- A low-resistance ohmmeter that reads the micro-ohm contact resistance across a bolted joint
FAQ
What megger voltage do I use on busway?
Use the test voltage from the manufacturer's published data or the NETA table of test values for the system rating. For busway rated 600 V and below a 1000 Vdc megger test is common, and medium-voltage bus is tested higher, often in the 2500 to 5000 Vdc range. Confirm against the manufacturer and the NETA acceptance edition.
Why is my busway megger reading low?
A low busway reading is usually moisture, not damage, because the insulation is hygroscopic and absorbs water in damp storage or when cold bus sweats in a warm room. Cold gear, a dirty surface, or a reading below the dew point all read low too. Dry the run and re-megger before condemning it.
How do I dry out busway that reads low?
Confirm the low reading is moisture, then warm the run to drive the water out, using a warm dry space, external blowers or heat lamps, the manufacturer's drying procedure, or a controlled low current through the bars. Watch the megger; the reading often dips then climbs and levels off when dry. Never energize wet busway to dry it.
What torque do busway joints take?
Torque busway joints to the manufacturer's published value, not a borrowed number. Most modern busway uses a single bolt with a twist-off head or visual indicator that shows when it is correct, with a common value around 70 lb-ft. When the head shears or the indicator shows, the joint is made. Mark each joint after you verify it.
What insulation resistance is acceptable for busway?
Acceptable busway insulation resistance is the manufacturer's published minimum and the NETA acceptance value, which scales to the run length, read against a dry baseline. A clean dry low-voltage run reads in the high megohms to gigohms. A reading in the low megohms on a dry run is a flag, so dry it, re-read, and record the corrected baseline.
Feeder busway vs plug-in busway: what is the difference?
Feeder busway has no tap openings and moves power point to point with the lowest drop and fewest joints. Plug-in busway has tap openings along its length at a fixed spacing for plug-in units, so it runs down the rack rows and lets you land power where the racks are. Read the type off the submittal.
Do I megger a plug-in unit before energizing it?
Yes. Megger the plug-in unit and its branch phase-to-phase and phase-to-ground before it feeds the rack PDU, the same way you megger the run. Also confirm the stabs engage the bus fully and squarely, the interlock works and is not defeated, and the overcurrent rating and phasing into the PDU are correct.
What do I do if a busway joint is not torqued or runs hot?
A joint that was not verified at install, or that scans hot on the infrared survey under load, gets shut down, opened, the contact surface cleaned, and the joint remade to the manufacturer's torque, then rescanned. A loose or oxidized joint runs hot and fails, so it is corrected before energization or scheduled for shutdown, not left in service.
Why does busway use bolted joints instead of welded ones?
Busway uses bolted joints so the run can be shipped in lengths, installed in place, and taken apart for changes and maintenance, which is the flexibility data centers buy busway for. The cost is that every joint is a potential hot spot, so the bolted joint is exactly where the torque verification, the ductor check, and the infrared scan focus.
Why does the busway housing matter for grounding?
On most busway the metal housing is the equipment grounding conductor, so its continuity across every joint is the fault-current path that lets the breaker clear a ground fault. A joint that is mechanically together but poorly bonded leaves that path missing when a fault needs it, so verify ground-bond continuity across the joints before energizing.
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.