Port Container Terminal Transformer: How to Size the MV Distribution for STS and RMG Cranes, Why Crane Duty Cycle and Regeneration Decide the kVA, and What a Coastal Terminal Demands

The answer first: a container terminal transformer is sized from a crane fleet load list with duty-cycle demand factors, and the crane is not a normal motor load. Ship-to-shore (STS) and rail-mounted gantry (RMG) cranes draw a cyclic load that swings between regeneration and 150% of average within a 90-second cycle, so the transformer is selected on thermal RMS and hot-spot temperature rather than on the sum of motor nameplates. The reefer plugs are the opposite problem — a steady, weather-driven load that sets the base of the curve and the summer peak. Size the MV substation on coincidence between those two families, then choose the crane transformers on cyclic rating, K-factor and voltage drop along the quay, and the kVA, the impedance, the IP rating and the dry-type-versus-oil decision all become calculated outputs instead of habits.

Key takeaways

• Size STS crane capacity at 1,200–1,600 kVA installed per ULCV twin-lift crane and apply a duty-cycle demand factor of 0.40–0.50 — not the 0.7–0.9 used for building loads. The coincidence between cranes on the same quay is high; the coincidence between cranes and reefers is not.
• Reefers are the largest steady load on a transshipment terminal. A 40-ft integral reefer draws roughly 5–9 kW steady and 11–13 kW during pull-down, and the plug itself is rated about 22 kVA at 32 A/400 V three-phase, so design demand is taken on kW at the design ambient, then converted to kVA at 0.85–0.90 power factor.
• Select crane transformers on the 10-minute RMS current and the hot-spot limit (98 °C for normal cyclic life, per IEC 60076-7), not on the instantaneous peak. A cast-resin unit with a 30–60 minute thermal time constant barely responds to a 20-second hoist peak, which is exactly why a 1,600 kVA unit can serve a crane whose instantaneous demand touches 2,000 kVA.
• Crane drives are the harmonic source that decides the transformer design. A 6-pulse thyristor or diode front end produces 25–35% current THD; 12-pulse halves it; an active front end (AFE) with regeneration brings it to 4–8%. Specify K-factor or a IEEE C57.110 derating, because a standard distribution transformer carrying K-13 harmonic content loses 10–20% of its usable capacity to eddy-current heating.
• Feed cranes at 6.6, 11 or 20 kV through a cable reel or conductor rail and transform on board or in a quay substation. At 400 V a 1,400 kVA crane at the end of a 400 m quay run drops about 9% at full load; at 6.6 kV the same load drops 0.3%. That single number decides the voltage.
• Coastal terminals are a C5-M corrosion environment with 40–45 °C ambient, salt fog and condensation cycling. That means IP54 minimum (IP55 on the crane), ISO 12944 C5-M paint systems, silicone or HV cast-resin insulators, stainless hardware, and a cooling design that assumes the enclosure is not a clean room.
• N-1 firm capacity at a terminal is not optional: a berth without cranes stops earning. The usual arrangement is a sectionalised 11 kV crane bus fed from two or three 33/11 kV transformers, so any single transformer outage leaves every berth with reduced crane count rather than one berth dead.

What loads does a container terminal have that an ordinary factory does not?

A factory has a process that sets its load profile. A terminal has a fleet, and the fleet moves. Four load families behave completely differently, and the transformer specification is decided by whichever one is worst for the parameter being specified — cranes for cyclic thermal duty and harmonics, reefers for steady kVA, cold ironing for low power factor at low load, and the automation and IT layer for outage tolerance.

The first thing this table shows is that an "installed kVA" figure is close to meaningless at a terminal. Summing 12 STS cranes at 1,400 kVA gives 16,800 kVA of connected capacity, and a naive design that applies a 0.8 factor would specify a 13 MVA substation for cranes alone. The cranes will never draw that: the hoist, trolley and gantry drives are interlocked by the crane's own control system, and the average demand of a working ULCV crane is 400–800 kVA with peaks that are short.

How do you build the crane load list and size the terminal substation?

The method has three inputs that are genuinely different from a building load calculation: the crane's installed kVA, its measured or specified duty cycle, and the berth productivity the terminal is designed to achieve. Everything else is arithmetic.

Step 1 — take installed kVA per crane from the drive schedule, not from the motor sum

A modern STS crane has a hoist, a trolley, a gantry (long travel) and usually a boom hoist, plus auxiliary loads for the spreader, lights, air conditioning and the machine-room ventilation. On a twin-lift ULCV crane the installed drive power typically lands between 1,200 and 1,600 kVA, but the figure that matters is the crane's own main transformer or main breaker rating, because the crane designer has already applied the simultaneity factors inside the crane. Ask for the crane's single-line diagram and the main incoming breaker rating; do not add motor nameplates.

That incoming rating also tells you the inrush the transformer will see. A crane with an AFE front end energises gently and draws little more than rated current; a crane with a 6-pulse diode front end and a large DC bus charges its capacitor bank through a pre-charge resistor, but the first movement of a cold hoist still produces a current step of 1.5–2.0 times rated. The inrush current guide covers the transformer side of this; on a terminal the generator or the upstream protection must be checked against the crane's starting step as well.

Step 2 — apply a duty-cycle demand factor, and a berth coincidence factor on top

Crane demand factors of 0.40–0.50 on installed kVA are normal for a working fleet, because a crane spends part of every cycle waiting, and the drives are not all at full power at once. Berth coincidence is higher than a factory's machine coincidence: when a vessel is worked, every crane on that berth works, and when the ship is alongside the whole quay is productive. The coincidence between berths is close to 1.0 during a surge.

The exception is the reefer load, which is anti-correlated with nothing and simply persists. Reefers do not care whether a vessel is being worked, and their demand peaks in the afternoon at high ambient, which on a transshipment terminal is also when crane productivity peaks because the yard is trying to clear the backlog.

Step 3 — build the coincident demand and then apply N-1 on top of it

A worked example for a 1.2 million TEU terminal with six berths, 12 STS cranes, 24 RMG, 1,200 reefer plugs, 2 cold-ironing berths and an electrified yard:

23 MVA of coincident demand lands on a 33/11 kV substation of two 20 MVA transformers, each loaded at about 58% at peak, with a third 20 MVA unit — or a pre-engineered mobile-substation bay — providing N-1 firm capacity. Two units alone would give only 20 MVA of firm capacity, which is below the design peak, and at a terminal that means shedding cranes on the busiest day of the year. The distribution system sizing guide covers the feeder level of the same calculation.

Why does the crane duty cycle, and not the peak kW, decide the transformer rating?

A transformer is a thermal device. Its life is consumed by hot-spot temperature, which follows the winding temperature with a time constant of minutes (cast resin) to a couple of hours (oil-immersed with a large oil mass). A crane cycle of 90 seconds therefore looks, to the transformer, like a slightly elevated steady load rather than a series of peaks. That is why the correct input to the rating calculation is the RMS current over a 10-minute window, plus a check that the hot spot stays within limits during the peak.

Put numbers on it. A crane with 1,400 kVA installed draws 700 kVA for 70 s of the cycle and 1,050 kVA for 20 s, and regenerates 450 kVA for 20 s. Over a 110 s cycle the RMS is about 700 kVA, so a 1,000 kVA transformer would look adequate on a nameplate basis — but the peak of 1,050 kVA at 0.8 power factor is 1.31 times nominal current, and the winding must stay below its hot-spot limit through the whole shift, not just through the peak. The practical outcome is that crane transformers are specified at 1.15–1.3 times the 10-minute RMS, which for this example is 800–900 kVA of transformer for a 1,400 kVA crane. That is the number that should appear on the datasheet, with the crane's duty class and the RMS calculation attached.

The duty class itself should be stated the same way the crane is classified. Cranes are classified by the number of cycles and the load spectrum (ISO 4301-1 groups A1 to A8 for the crane as a whole, FEM 9.511 for mechanisms, M1–M8), and a terminal crane working 24/7 sits in the top groups. A transformer bought for an M4 duty and installed on an M8 crane will run hot for its whole life, and no amount of derating after the fact fixes the winding time constant. Put the crane's duty group on the enquiry.

Why do STS cranes inject more harmonics than a normal factory?

A factory has one large drive repeating the same motion. A crane has four drives, each accelerating and decelerating independently, drawing power through a rectifier that is only partially loaded most of the time. The harmonic spectrum therefore changes continuously, and at light load the current THD is worst — a drive that draws 8% THD at full load can draw 40% at 20% load, because the fundamental collapses while the commutation notches stay the same size.

The 12-pulse case is worth understanding because it moves the harmonic problem into the transformer. Two 6-pulse bridges fed from a delta and a star secondary, with 30° of phase displacement between them, cancel the 5th and 7th harmonics in the primary. That cancellation only holds if the two secondaries are loaded equally, which they are not when one trolley drive is idle, so a 12-pulse crane still needs a transformer designed for the residual spectrum, not for an ideal textbook cancellation. A three-winding transformer built for this duty is a different product from a two-winding distribution unit, and it should be specified as such.

For any of these cases, K-factor is the language to specify in. K-factor weights each harmonic by the square of its order, because eddy-current loss rises with the square of frequency, and a K-13 transformer is designed to carry a spectrum whose K-factor is 13 without exceeding its rated hot spot. The K-factor selection guide gives the calculation, and the harmonics guide covers the physics and the mitigation options. On the utility side of the point of common coupling, IEEE 519-2014 sets the current distortion limits; at a terminal the PCC is usually the 33 kV incoming, and the crane harmonics are attenuated by the 33/11 kV transformer impedance, which is one more reason not to make that transformer's uk% unnecessarily high.

Why is the crane fed at 6.6–11 kV instead of 400 V?

Because the quay is long. A crane travels the full length of a berth, and the cable reel carries 300–600 m of cable. At low voltage that cable's resistance and reactance dominate the design; at medium voltage they are irrelevant.

The voltage drop calculation is the standard one:

ΔU% = √3 × I × (R cosφ + X sinφ) ÷ U_n × 100

Take a 1,400 kVA crane at 0.9 power factor, fed over 400 m:

Nine percent at 400 V is not a design margin problem, it is a failure: the crane's drives will trip on undervoltage during a hoist start, and the fix is either a cable so large that the reel cannot carry it or a medium-voltage distribution. This is why STS cranes are fed at 6.6 or 11 kV through a motorised cable reel or a conductor rail in the quay, with the transformer either on board the crane (most common, 11 kV or 6.6 kV to 0.69 kV or 0.48 kV) or in a quay substation feeding an LV busbar. The voltage regulation guide covers the calculation in more detail, including the effect of power factor on the resistive and reactive terms.

The choice between 6.6 and 11 kV is usually a site standard rather than a technical one, but two things do change: at 11 kV the cable and reel are smaller and cheaper for the same power, and the transformer's insulation level and BIL rise, which costs a little more. Where the terminal also has cold ironing at 6.6 or 11 kV, aligning the crane voltage with the shore-power voltage lets the same switchgear family and spares serve both. Terminals in 60 Hz networks face the additional issue that a 60 Hz crane bought for one project may be redeployed to a 50 Hz terminal; the 50 Hz versus 60 Hz guide explains why the V/f ratio, not the nameplate frequency, decides whether that works.

What does regenerative braking do to the transformer and to the energy bill?

Every lowering move returns energy. A hoist lowering a 40-tonne spreader load recovers 30–45% of the energy the hoist consumed lifting it, and a gantry decelerating recovers its kinetic energy. On a crane with a diode front end that energy is burned in a braking resistor and appears as heat in the machine room; on a crane with an AFE it is pushed back into the 11 kV bus, where it can be absorbed by another crane that is hoisting, by the reefers, or exported to the utility.

Three consequences for the transformer. First, the transformer must be able to carry reverse power, which is a protection and metering question rather than a thermal one but is often overlooked when a diode-front-end crane is retrofitted with AFE drives. Second, the harmonic spectrum in regeneration is not the same as in motoring, so a filter tuned for motoring may not cover the regenerating case. Third, the energy that is recovered has to go somewhere: if the terminal has no other load, the 33/11 kV transformer carries it back to the utility, and some utilities charge for reactive export or require a reverse-power relay. Where the regenerated energy is large and the utility connection is weak, terminals install a flywheel or battery storage at the 11 kV bus, which is a separate asset but changes the transformer's RMS loading and should be in the load study.

Should the crane transformers be dry-type or oil-immersed?

The answer splits by location, not by preference. On board a crane, the transformer sits in a machine room above the quay, often with limited ventilation, and it is a fire risk that the port authority will not accept in oil form. In a quay substation or the main 33/11 kV substation, oil-immersed units are cheaper per kVA, tolerate the coastal ambient better and have a longer thermal time constant, which suits cyclic loading.

The on-board case has a detail that catches buyers: the crane manufacturer often specifies the transformer as part of the crane package, and the terminal's own specification never reaches it. If the port wants a particular insulation class, a particular IP rating or a K-factor, it has to be written into the crane purchase order, not into the transformer order that arrives later. The oil versus dry-type guide covers the general selection logic, and the cooling methods guide explains why an AN/AF cast-resin unit can be rated at 1.25–1.5 times its AN rating for the short crane peak, which is often the cheapest kVA a terminal can buy.

What IP rating and corrosion class does a coastal terminal need?

A container terminal is one of the harshest electrical environments in normal industrial use: salt-laden air, 80–95% relative humidity, condensation every night when the steel cools below dew point, wind-blown sand, and a 40–45 °C ambient in the tropics with solar gain on top. IP54 is the minimum for any transformer enclosure at a terminal, and IP55 is normal on the crane because of washdown and driving rain. IP rating protects against water and dust; it says nothing about corrosion, which is a separate specification.

The corrosion specification should be ISO 12944-2 category C5-M (very high corrosivity, marine) for anything within a few kilometres of the water, with a paint system of at least 240 µm total dry film thickness in a zinc-rich primer, epoxy intermediate and polyurethane topcoat. Practical details that fail first on real terminals: zinc-plated fasteners (use stainless 316), unpainted tank seams, mild-steel cable trays, standard breathers that ingest salt mist, and rubber gaskets that harden in UV. For oil-immersed units, a hermetic design or a conservator with a membrane bag avoids breathing moist coastal air altogether.

Ambient temperature is the second half of the same problem. A transformer rated at 40 °C ambient loses roughly 8–10% of its capacity at 45 °C, and a cast-resin unit in a steel enclosure in the sun can see 50 °C internal air. The derating must be applied to the crane's RMS figure, not discovered later, and the enclosure ventilation must be designed for the actual heat rejection, not for the transformer's nameplate loss at 20 °C.

How do you keep the cranes running when a transformer fails?

Terminal availability is measured in moves per hour, and a transformer outage that stops two cranes for a week costs more than the transformer. The design answer is a sectionalised MV bus with N-1 capacity, plus a spare strategy that is agreed before the first unit is ordered.

The common arrangement is a double busbar or a single busbar with a normally-open section at the 11 kV level, each half fed by its own 33/11 kV transformer, and each quay crane fed from a busbar section through a cable reel. When one transformer is out, the section breaker closes and the remaining transformer carries the full quay at reduced crane count, which is a productivity loss rather than a stoppage. Where the terminal cannot accept even that, the third transformer is energised and on hot standby, which costs no-load loss every hour of the year but removes the switching time.

Protection has to match the topology. Crane feeders see repeated starting current, regeneration, and a cable reel that is a moving mechanical component with a finite life — the earth-fault protection on a reel-fed crane must be sensitive enough to detect a fault at the far end of 600 m of cable while not tripping on the normal capacitance of that cable. Transformer differential protection on a 20 MVA unit is standard; on a 1,600 kVA on-board crane transformer it is usually replaced by restricted earth fault and a fuse or breaker on the MV side, because the crane's own protection is coordinated with the port's. The protection devices guide covers the device set, and the short-circuit calculation guide gives the fault level method that the settings depend on.

How does impedance change the fault level and the crane's voltage dip?

Two requirements pull in opposite directions. A low uk% keeps the voltage dip during a hoist start small and helps the crane ride through; a high uk% limits the fault level at the 0.4 kV or 0.69 kV busbar so that the crane's breakers and busbars can be standard products. At a terminal the dip is usually the binding constraint because every crane start is visible on the same bus.

The dip calculation is the same as the voltage-drop calculation at starting current. A 1,600 kVA transformer at 5.5% uk carrying a crane that steps to 2.5 times its rated current for one second will dip the bus by roughly 5.5% × 2.5 ≈ 13.75% if the upstream source were infinite, and by more once the 33/11 kV transformer and the utility source impedance are added. Most crane drives tolerate 15–20% dip for a second; some AFE drives trip at 10% below nominal for more than a few cycles. That tolerance is the number to design against, and it should be obtained from the crane supplier in writing rather than assumed.

What belongs in the enquiry and on the datasheet?

Two terminal transformer offers are comparable only if both answer the same questions. The list below is short enough to send with an RFQ and specific enough that a supplier cannot answer it with a standard catalogue sheet.

The two fields most often missing from real enquiries are the duty class and the harmonic spectrum, and they are the two that change the price most. A supplier who is not told the crane duty will quote a standard distribution transformer at standard losses, and the terminal will discover the difference in the third summer. Send the crane's drive schedule, the single-line diagram, the reefer plug count and the design ambient together, and ask for the loss figures and the derating calculation in the offer. The pre-shipment inspection checklist covers what to verify before the unit ships, which at a terminal matters because the crane cannot be commissioned without it.

How does the transformer choice affect twenty-year cost?

A terminal operates 6,000–8,000 hours a year, and the losses are paid at industrial tariffs that are frequently above USD 0.10/kWh. Take a 1,600 kVA on-board crane transformer with 2.4 kW of no-load loss and 16 kW of load loss at full load, averaging 45% loading over 7,000 hours:

Loss = 2.4 + 16 × 0.45² = 2.4 + 3.24 = 5.64 kW, which over 7,000 hours is 39,480 kWh a year. At USD 0.10/kWh that is USD 3,948 a year, and the present value over twenty years at 6% is about USD 45,300. Across 12 STS cranes the loss bill is more than half a million dollars of present value, which is usually several times the price difference between a standard and a low-loss design.

That arithmetic argues for two things: specify losses and capitalise them in the evaluation, and do not oversize. A crane transformer running at 25% average load pays almost the same no-load loss as one at 50% but delivers half the moves, and terminal designers who add 30% "future capacity" to every crane transformer end up paying for it for two decades. The total cost of ownership guide gives the capitalisation method, and the efficiency standards comparison gives the regulatory minimums that apply in the EU, the US and China — at a terminal those minimums are a floor, not a target.

If you are preparing an enquiry for a terminal, a quay substation or a crane package, the fastest route to a comparable offer is to send the crane fleet list with installed kVA and duty class, the reefer plug count and design ambient, the single-line diagram with the proposed voltages, and the harmonic study or drive schedule. Our engineering team can turn that into a sized transformer specification with the cyclic rating, the K-factor derating, the voltage-drop check and an indicative budget in one pass. You can also start from the power system calculators to check the fault level, the voltage dip and the loss capitalisation before you approach any supplier.

Sources / 资料来源

Source: IEC 60076-1 — Power transformers, general requirements.
Source: IEC 60076-2 — Temperature rise for liquid-immersed transformers.
Source: IEC 60076-3 — Insulation levels, dielectric tests and external clearances in air.
Source: IEC 60076-5 — Ability to withstand short circuit.
Source: IEC 60076-7 — Loading guide for oil-immersed power transformers, including cyclic loading and hot-spot limits.
Source: IEC 60076-10 — Determination of sound levels.
Source: IEC 60076-11 — Dry-type power transformers, including climatic, environmental and fire behaviour classes.
Source: IEC 60137 — Insulated bushings for alternating voltages above 1,000 V.
Source: IEC 60214 — Tap changers.
Source: IEC 60034-1 — Rotating electrical machines, rating and performance, duty cycles S1–S10.
Source: IEC 61439-1 and IEC 61439-2 — Low-voltage switchgear and controlgear assemblies.
Source: IEC 61936-1 — Power installations exceeding 1 kV AC, common rules.
Source: IEC 60529 — Degrees of protection provided by enclosures (IP code).
Source: ISO 4301-1 — Cranes, classification, general (crane duty groups A1–A8).
Source: FEM 9.511 — Rules for the design of series lifting equipment, classification of mechanisms.
Source: IEEE C57.110-2018 — Establishing transformer capability when supplying non-sinusoidal load currents.
Source: IEEE 519-2014 — Harmonic control in electric power systems.
Source: IEEE C57.12.00 — General requirements for liquid-immersed distribution, power and regulating transformers.
Source: IEEE C57.91 — Guide for loading mineral-oil-immersed transformers and step-voltage regulators.
Source: GB/T 6451-2015 — Technical parameters and requirements for three-phase oil-immersed power transformers.
Source: GB 20052-2020 — Minimum allowable values of energy efficiency and energy efficiency grades for power transformers.
Source: Commission Regulation (EU) No 548/2014 — Ecodesign requirements for small, medium and large power transformers.
Source: U.S. Department of Energy, 10 CFR Part 431 Subpart K — Distribution transformer energy conservation standards.
Source: ISO 12944-2 — Corrosivity categories for steel structures, category C5-M for coastal atmospheres.
Source: NFPA 70 (National Electrical Code) Article 450 — Transformers and transformer vaults.
Source: QDTB engineering experience — container terminal crane load lists, STS crane duty-cycle RMS sizing, reefer plug demand factors, quay voltage-drop checks, and harmonic derating for crane drives. Values marked as industry experience are typical ranges from port and heavy-lift projects, not published standards.