A gantry crane travels further than any machine in the plant, and the power has to follow it. Transformer, cable, reel and earth, measured on the runway rather than on the nameplate.
A combination crane system is not one crane. It is a runway with several bridges and hoists hanging off a shared conductor line, and the electrical design comes down to two numbers that rarely appear on the drawing: the current at the feed point, and the distance from that feed to the farthest collector shoe. Get those right and the rest is fittings. Get them wrong and the station at the far end trips on undervoltage every time it lifts.
We build these systems from 0.125 to 2 tonnes per hoist, on tracks that are straight here and switchable there. This article is about the part that ships as a box of bars, hangers and shoes, and about the three places where a shared supply behaves differently from six separate ones.
Because the crane is the only thing on the quotation. Each bridge arrives with its own panel, its own contactors and its own overload protection, and each one is tested on a bench before it leaves. The shared part is upstream: one feed, one run of conductor bar, one collector set per bridge. Nothing on that chain carries a serial number that says "crane", so nobody puts it on the drawing.
The symptoms then show up as crane faults. A hoist that refuses to lift the last 20 percent of its rated load. A contactor that chatters when a second bridge starts. A brake that releases a fraction late. A drive that trips on undervoltage during acceleration and resets fine. All of those read like a bad hoist and none of them are.
There is a simple way to tell the difference. Note where on the runway it happens. If the fault follows the position and not the machine, the problem is in the distribution and the crane is innocent.
Five different limits get confused with each other, and only one of them is electrical.
| Limit | Set by | What it controls |
|---|---|---|
| Number of bridges the runway can hold | Track and support load calculation | Structural capacity, with buffers or intermediate bumpers keeping two bridges out of the same support bay |
| Current the bar can carry | Bar ampacity at the duty in use | Sum of the currents of every bridge working at the same time |
| How far the power reaches | 3 percent voltage drop, CMAA practice | Distance from the feed to the farthest collector shoe, not total runway length |
| Whether control reaches the trolley | Control architecture | A five pole bar is three phases, neutral and earth. Nothing is left for control |
| Whether power crosses a switch | Switch design | Some switch types cannot be electrified at all |
The structural side is well understood by the people who sell you the track. A 2,000 lb runway can carry two 1,000 lb bridges or four 500 lb bridges, and the mechanism that makes it legal is a buffer that rolls in the runway between two bridges so they cannot both sit in the same span. NIKO states the same rule in metric terms: two 500 kg bridges are fine on 500 kg runways as long as they are prevented from operating inside the same two consecutive support points.
What that arithmetic does not tell you is how the runway is wired. Four bridges on a runway sized for four bridges still need a bar sized for four bridges, and the bar is chosen on a different calculation, by a different person, usually later. Our combination crane order sheet asks for the number of bridges and the track layout together for exactly this reason.
The rule itself is short. CMAA practice, repeated in every conductor bar catalogue, is that voltage drop from the power tap to the unit motors at the farthest point on the conductor run shall not exceed 3 percent. For three phase work the arithmetic is drop equals length times current times impedance times 1.73, with length in the units the impedance is quoted in.
The part that gets lost is that a combination crane has no fixed far end. Bridges move. So the honest design case is the worst arrangement the layout allows, not the average one.
If the bridges are evenly spread along the run, the current in each segment falls as you move away from the feed, and the drop to the last bridge works out to the lumped figure multiplied by (n+1) divided by 2n. For six bridges that factor is 0.583. For four it is 0.625. It never reaches 0.5 in practice.
That factor is real, but it assumes even spacing, and even spacing is the one thing an operator will not maintain. Buffers and intermediate bumpers help. Without them, all six bridges can end up in the last ten metres, and the factor goes back towards 1.0. We design buffered lines at 0.75 and unbuffered lines at 1.0, and we write the assumption on the drawing so the next person can see what it depends on.
| Feed arrangement | Length to farthest shoe | Drop at 40.8 A | Percent of 400 V |
|---|---|---|---|
| One feed, one end, no buffers | 60 m | 15.3 V | 3.83 percent, fails |
| One feed, one end, six bridges buffered | 60 m | 11.5 V | 2.87 percent, passes by 0.13 |
| One feed at the centre | 30 m | 5.7 V | 1.44 percent |
| Two feeds, one at each end | 30 m, half the current each | 2.9 V | 0.72 percent |
These are our own calculations for a galvanised steel figure eight bar with an impedance of 0.0011 ohms per foot, on a 60 m runway carrying 40.8 A, at 400 V. They are planning figures, not catalogue data. Two conclusions survive any reasonable change to the inputs. A centre feed halves the drop for the price of one extra cable. A second feed at the far end quarters it. Neither is expensive compared with the copper that would otherwise be needed in the bar.
Here is where the catalogue tries to help and mostly confuses. Ampacity and voltage drop are two different questions, and the bar that wins on the first loses on the second.
| Bar | Impedance | Rated current | Drop at 60 m | Longest run at 3 percent |
|---|---|---|---|---|
| Galvanised steel figure eight, 90 A | 0.00361 ohm/m | 90 A continuous, 135 A intermittent | 2.87 percent buffered, 3.83 unbuffered | 63 m buffered, 47 m unbuffered |
| Galvanised steel figure eight, 110 A | 0.00262 ohm/m | 110 A continuous, 165 A intermittent | 2.08 percent buffered | 86 m buffered, 65 m unbuffered |
| Stainless steel with copper laminate, 160 A | 0.000472 ohm/m | 160 A continuous, 240 A intermittent | 0.38 percent buffered | 480 m buffered |
| Rolled copper, 350 A | 0.000279 ohm/m | 350 A continuous, 530 A intermittent | 0.22 percent buffered | 812 m buffered |
| Compact copper bar, light section | 0.00054 ohm/m at 50 Hz | 285 A at 60 percent duty, 220 A at 100 percent | 0.43 percent buffered | 419 m buffered |
Two things to notice. First, the 90 A bar is electrically marginal at 60 m even though it is 2.2 times the current we are asking it to carry. Ampacity answers a thermal question. It says nothing about length.
Second, the same bar family quotes 285 A at 60 percent duty and 220 A at 100 percent duty. That is a 23 percent swing from duty alone, before ambient temperature. A crane runway is closer to continuous than a machine tool, so read the 100 percent column, not the headline number.
If you extend a modular line, which is the whole point of buying one, remember that the bar you bought for 60 m is at 2.87 percent. Every metre you add takes you further past the limit. That is a design failure built into the expansion the salesman described as free.
A track switch is the reason to use conductor bar in the first place. A festoon needs a continuous cable path, and a switch breaks it. Most manufacturers say this plainly: conductor bars are the recommended solution for runways that use a curve or a track switch, and a monorail with a complex curve or switch cannot use a standard C track or I beam for power delivery.
What the same catalogues say less loudly is that not every switch can be electrified. One widely used sliding monorail switch cannot be electrified at all, in the manufacturer's own words, because of space limitations.
| Junction | Can it be electrified | What it needs |
|---|---|---|
| Straight rail splice | Yes | Splice joint plus a jumper across the expansion gap |
| Sliding monorail switch, space limited design | No | Nothing. The track section is too shallow for bar plus collector |
| Rotating crossover, right angle | Yes, 2 to 4 conductors | Conductors at 156 mm above the tread, wiring harness acting as a jumper to both incoming tracks |
| Curved section | Yes | Segmented bar built to the radius, with the collector shoe rated for the same curve |
| Interlocked transfer between two cranes | Yes, if the harness is supplied | Interlocking system per ASME B30.17, with power continuity passing through the interlock rather than through the track |
At a sliding switch the moving section carries its own short conductor segment, and that segment is only live when the switch is seated against the incoming track. The collector shoes cross a gap in the bar. If the gap is longer than the distance between the first and last shoe on one bridge, the bridge loses power in the middle of the transition. On a hoist that is holding a load, losing power means the brake drops.
This is why the shoe spacing on a bridged collector set matters, and why a cheap collector set with two closely spaced shoes can pass every bench test and fail on the one switch in the layout. Ask for the gap length and the shoe spacing as numbers before the drawing is approved. If the switch supplier does not have the gap length, the layout is not ready.
One more constraint that surprises people: guarded runway conductors are now a code requirement in the United States, and open conductors are no longer recommended for new work. Older plants run bare bars, and that is fine until someone adds a bridge or extends the track and finds the insurance surveyor reading the same clause.
A five pole conductor bar is three phases, a neutral and a protective conductor. It is full. NIKO states the consequence in one line in its own catalogue: conductor bars are only suitable for power supply, not control supply. Demag's integrated conductor line is offered for two profile sizes and is five pole for the same reason.
That leaves the control circuit with nowhere to ride, and on a switched track the pendant cable cannot follow the bridge either.
| Control option | Works through a switch | Typical cost | How it fails |
|---|---|---|---|
| Pendant station on the bridge, cable along the bridge only | Yes | Lowest, included | Cable wear where it leaves the bridge end |
| Radio remote | Yes | USD 600 to 1,500 per bridge | Battery, interference, and no tactile feedback |
| Separate control festoon in its own box track | No | USD 25 to 45 per metre | Stack up at the travel ends, cable fatigue |
| Extra poles in the conductor bar, where the family offers them | Yes | 30 to 40 percent more bar | Pole spacing, and 110 V control next to 400 V power |
For a line with one switch, radio is usually the answer and it is usually priced after the contract is signed. That is the wrong order. A switched track changes the control architecture, and the control architecture is not a detail you discover at commissioning.
The protective pole in a conductor bar runs continuously along fixed track, and each bridge collects it with a dedicated shoe. At a moving section it breaks. The moving track then needs its own bond to the fixed structure, and that bond has to be a conductor, not a bearing.
The relevant principle is old and still widely ignored: neither the trolley frame nor the bridge frame may be considered earthed through its wheels and the track. A rolling contact is not a bonding conductor, and neither is a pivot pin or a set of rollers under a rotating crossover.
| Point | How to bond it | What to measure |
|---|---|---|
| Rotating crossover or turntable | Flexible copper conductor from the moving frame to the fixed frame, sized for the fault current | Resistance across the joint, below 0.1 ohm |
| Sliding switch | Bond through the latch frame or a trailing conductor, never through the sliding surfaces | Continuity in every switch position, not just the seated one |
| Bridge to runway | Earth shoe in the conductor bar, plus the frame bond | Resistance from bridge frame to the building earth bar, below 0.1 ohm |
| Whole installation | Earth electrode and structure bonding | Earth resistance, 4 ohm or less |
We check this at handover with a low resistance meter reading at least 10 A, because a multimeter will happily show 0.3 ohm through a painted joint and call it bonded. The same inspection round measures insulation of the bar at 500 V and expects at least 0.5 megohm to earth.
Two different discount factors get applied to two different parts of the same system, and mixing them up is how a bar ends up undersized.
The first is per unit. Within one crane, the hoist counts at 100 percent, the trolley at 50 percent and the bridge travel at 50 percent, because a crane does not accelerate all three motions at full power at once. The second is per installation. Across several cranes on one runway, further demand factors apply, and the published example for two cranes is 0.95.
| Basis | Power | Current at 400 V | Use it for |
|---|---|---|---|
| Six stations, installed | 25.2 kW | 47.5 A | Nothing. It is the sum of nameplates |
| Per unit factored, 3.0 + 0.2 + 0.4 kW each | 21.6 kW | 40.8 A | The conductor bar and the collector shoes |
| Multi unit demand at 0.75 | 16.2 kW | 30.6 A | Incoming feeder, main breaker, isolator |
The bar is sized on the per unit figure, not the discounted one. The bar is the thing the current physically flows through on its way to every bridge, so it sees the sum, not a statistical average of the sum. The incoming cable and the main breaker can use the discounted figure because they see the line as a whole.
The 0.75 in the third row is ours, not a published table. Published multi crane tables get thin above two units, and any figure below 1.0 assumes that six operators do not all lift at the same moment. Buffers help. Interlocks help more. If your line has neither, size the feeder at the per unit figure too and accept a slightly larger cable.
A single feed is the cheapest and safest arrangement for a line that runs as one process. One isolator, one key, and everything downstream is dead. The problem is the other six days of the week, when you need to change a collector shoe on bridge four and the other five bridges are supposed to keep working.
This is a known failure of a good procedure. Lockout works beautifully on one crane on one runway and gets much harder on several cranes sharing one conductor run. In practice, when isolating the line stops production, people stop isolating the line.
| Arrangement | What it isolates | Cost |
|---|---|---|
| Single feed, one lockable isolator | The whole line | USD 250 to 600 |
| Sectionalised bar, lockable isolator per section | Part of the runway, while the rest stays live | USD 400 to 900 per extra section |
| De-energised maintenance section at the track end | A short length where a bridge can be parked for collector work | Designed in, small |
| Lockable plug and socket per bridge | The bridge, but the conductor bar above it stays live | USD 120 to 350 per bridge |
Read the third column of that last row carefully. Locking out a bridge does not de-energise the bar it hangs from. The bar is a bare conductor at 400 V, a metre above the head of whoever is working on the hoist. Any procedure that calls that "isolated" is wrong, and it is the most common version of wrong that we see on existing lines.
The maintenance section at the end of travel exists precisely to solve this. Some suppliers build it as standard on gantry runways so collectors can be inspected with the crane parked and the local section dead. It costs almost nothing at the design stage and cannot be retrofitted cheaply.
For a 60 m runway with six 2 tonne stations and one switch, our own quotation ranges run as follows. These are SIEC planning figures in US dollars, ex works, 2026 prices, and they exist to show proportion rather than to be a price list.
| Item | Quantity | Unit | Total |
|---|---|---|---|
| Conductor bar, four pole, 60 A class | 60 m | USD 35 to 80 per m | USD 2,100 to 4,800 |
| Collector set and arms | 6 | USD 120 to 350 | USD 720 to 2,100 |
| Main isolator and feed connection | 1 | USD 250 to 600 | USD 250 to 600 |
| Switch electrification kit | 1 | USD 350 to 1,200 | USD 350 to 1,200 |
| Second feed with lockable isolator | 1 | USD 400 to 900 | USD 400 to 900 |
| Radio remote, required once the track has a switch | 6 | USD 600 to 1,500 | USD 3,600 to 9,000 |
| Distribution total | USD 7,420 to 18,600 |
Six 2 tonne combination stations, at a typical USD 4,500 to 9,000 each, come to USD 27,000 to 54,000. The distribution is therefore around a third of the hardware value, and it is the third that gets cut when a project runs over budget. It is also the third that determines whether the last station works.
If that sounds like a lot for some aluminium and copper, compare it with the alternative. Six separate feeds need six runs of cable, six isolators and six entries into the building distribution board, and they still cannot cross the switch.
The following distribution is from our own service records on modular systems, not from a published survey. It is here because the pattern is consistent enough to be useful when you are deciding what to inspect.
| First year issue | Share | Prevention |
|---|---|---|
| Collector shoe misaligned at a joint or switch | 28 percent | Align the bar to the track, not the track to the bar, and check shoe centring within 2 mm at each joint |
| Loose bar joint or overheated splice | 19 percent | Torque every splice at commissioning and again after 200 hours, then thermal scan annually |
| Control lost at the switch | 17 percent | Decide radio versus cable before the layout is frozen, not after |
| Drive undervoltage trips at the far end | 13 percent | Size the bar on length, add a centre feed, and measure the drop with a load bank at handover |
| No protective bond on a moving section | 9 percent | Bond the moving frame with a flexible conductor and measure it in every switch position |
| Carbon dust or moisture in the bar | 8 percent | Sealed bar in washdown or outdoor areas, and blow the holder out on the monthly round |
| Buffers removed, two bridges in one bay | 6 percent | Weld the intermediate bumpers, and put the rule in the operating instruction rather than on a sticker |
Notice what is not on the list: no failed hoist motor, no failed contactor, no failed bar. Every item is either an installation detail or a decision taken too late. That is a reasonable summary of electrical work on modular systems in general.
The inspection side is not complicated. Collector shoes get replaced at 60 to 70 percent of contact material consumed, brush thickness below 10 mm rather than at failure, spring pressure checked against the manufacturer figure (published values range from 15 to 30 N up to 36 to 67 N depending on the collector), and the shoe centred within 2 mm. Light duty bars under four hours a day last 18 to 24 months per brush set. Multi shift operations replace in 8 to 12 months. Waiting for a visibly worn brush means you are already damaging the bar.
Ten lines. If a supplier cannot answer them, the layout is not ready to be priced, and the gap will be discovered on site.
IEC 60204-32 covers all of this, and it is worth knowing how it words its own scope. The standard runs from the crane supply switch, includes power and control feeders outside the machine such as conductor bars, and explicitly applies to a group of hoisting machines working together in a coordinated manner. A combination crane system is not an edge case for that standard. It is the case the standard anticipated.
On the American side, ASME B30.17 covers cranes and monorails with underhung trolleys or bridges, and the 2026 revision includes the interlocking systems and drop sections that make multi track layouts legal. It was consolidated with B30.11 in 2015, which is why older drawings cite a standard number that no longer exists on its own.
Yes, and it is the main reason to choose bar over festoon on a shared runway. A single feed point energises the whole run, and each bridge collects from it with its own shoe set. NIKO puts the same point in its catalogue as an advantage of conductor bar over internal festoon, and Demag offers a five pole conductor line integrated into two of its aluminium profiles for exactly this case, including layouts with more than two cranes on one track. Festoon systems in light crane ranges are commonly limited to two bridges, because each bridge needs its own festoon and the parking area grows.
With conductor bar, and not with every switch. A festoon cannot follow a cable path through a switch, so bar is the usual answer. But at least one widely used sliding monorail switch cannot be electrified at all because there is no space for bar plus collector in the track section. Rotating crossovers can be electrified with two to four conductors, mounted about 156 mm above the track tread, with a wiring harness that acts as a jumper to both incoming tracks. Ask for the specific switch model and its electrification kit before you approve the layout.
Not normally. A five pole bar is three phases, neutral and protective earth, and that is the whole bar. NIKO states in its own catalogue that its conductor bars are suitable for power transmission only, not for control supply. Once the track has a switch, a pendant cable cannot follow the bridge either, so the realistic choices are a radio remote, a separate control festoon in its own track, or a bar family with additional poles. Decide this before the layout is fixed, because it changes the price of every bridge.
About 40.8 A after per unit factoring, which most 60 A bars will carry thermally without difficulty. The bar size that matters is the one that passes the 3 percent drop rule. A galvanised steel figure eight bar rated 90 A reaches 3.83 percent at 60 m with no buffers and 2.87 percent with buffers, so it passes by 0.13 of a percentage point and fails the moment the track is extended. Either specify the 110 A bar or move the feed to the centre, which halves the drop. Our combination crane datasheet carries the feed position as a field for this reason.
Not with a single feed. One isolator on a single feed de-energises the entire line, which is safe and stops production. The practical answer is to sectionalise the bar with a lockable isolator per section, and to design a short de-energised maintenance section at the end of travel where a bridge can be parked for collector work. What does not work is locking out one bridge and calling the bar isolated. The bridge is dead, the bar above it is still at 400 V.
Usually, yes, and on a switched track it is the only option that works at all. Six separate feeds need six cable runs, six isolators and six ways in the distribution board, and none of them can cross the switch. The trade is that a single feed concentrates the risk: one loose splice takes out six stations, and the current at that splice is the sum of everything downstream. That is a reason to specify the splice properly, not a reason to avoid the arrangement.
Related articles:
Gantry Crane Power Supply: What 200 Metres of Runway Does to Your Voltage, 5 to 100 Ton
Jib Crane Slip Ring: What Four Wires Have to Cross Before the Hoist Moves, 0.125 to 5 Ton
Traveling Jib Power Supply: The Three Moving Joints Nobody Draws, 0.125 to 2 Ton
Suspension Crane Power Quality: Why an Underhung Crane Is Really Part of the Building, 0.5 to 5 Ton
Crane Harmonic Distortion: What Five Drives on One Busbar Do to a Double Girder, 5 to 100 Ton