Electrical Systems

Every marine electrical design starts from a load list, and it stays the reference document for the vessel's life — every generator, cable and switchboard section traces back to a line on it. Each item carries a connected rating (its nameplate kW), a duty type (continuous, intermittent, or standby), and a load factor for each operating condition the ship runs in. The load factor captures the fact that a steering gear motor, say, does not run at full rated power continuously even while it is switched on — it cycles, and its average draw over the condition is a fraction of its nameplate figure.

The list is built and summed separately for each operating condition — sea-going, harbour/cargo work, manoeuvring, and emergency — because the mix of running equipment is different in each, and one condition will turn out to set the worst-case absorbed demand. It is that worst-case absorbed figure, not the sum of every nameplate rating on the ship, that generators are sized against; summing connected ratings assumes everything runs flat out simultaneously, which never happens and would produce a plant many times larger than the ship needs.

A well-built load list also records diversity between similar consumers (not every pump on a common header runs together) and leaves room for equipment added later in the ship's life, which is carried forward as an explicit growth margin rather than folded invisibly into individual load factors.

Once the worst-case absorbed demand and its growth margin are known, the next decision is how many generating sets to fit and what size each one should be. The usual philosophy is n+1: enough duty sets to cover the design demand between them, plus one further identical set on automatic stand-by, so that the failure or planned maintenance of any single generator still leaves the ship able to run essential services — with non-essential load shed automatically rather than left for an operator to notice and act on.

Sizing is not simply a matter of picking the smallest set that covers the demand. Diesel generators are designed to run best within a continuous loading band, typically from around half to a little over four-fifths of their rated output; run persistently lighter than that and unburnt fuel and carbon build up in the exhaust and cylinders (commonly called wet-stacking), while running at or above the rated output for long periods shortens the overhaul interval. The duty arrangement is chosen so the normal running load sits comfortably in that band, with enough spare capacity to start the largest motor on the board without an unacceptable voltage or frequency dip.

• Frequency and voltage matching — sets must be brought to the same frequency and voltage, and phase-synchronised, before their breakers can be closed together onto a common bus.
• Load sharing — once paralleled, governor and automatic voltage regulator droop settings make each set take a proportional share of the real and reactive load, so they support the bus together instead of one set hunting against another.
• Bus-tie and source interlocks — breaker interlocking prevents two independent sources (two generators, or a generator and shore supply) from being paralleled without the synchronising checks being satisfied.

Marine distribution is almost always radial: power flows outward from the main switchboard through distribution boards to final circuits, with no ring or mesh to complicate fault-finding. What survivability the arrangement has comes instead from splitting the main switchboard itself into two (or more) sections joined by a bus-tie breaker, each section fed by its own generators where practicable and each capable of carrying the essential load of the whole ship on its own if the tie is opened. A fault or fire confined to one switchboard section then costs half the plant, not all of it.

From the main board, feeders run to sub-distribution boards serving groups of consumers by area or function — engine room services, deck machinery, accommodation, navigation — rather than every final circuit being wired directly back to the main board. This keeps individual feeder lengths and fault levels manageable and lets a whole functional group be isolated for maintenance without touching unrelated services.

Within each board, circuits are explicitly split into essential and non-essential categories, because that split is what the automatic load-shedding scheme acts on when a generator is lost or overloaded: non-essential feeders (typically domestic and hotel loads, or non-critical HVAC) are tripped first and automatically, well before the remaining generating capacity is threatened. Voltage levels are chosen to suit plant size — low-voltage distribution (commonly in the few hundred volt three-phase range) covers most ships, with higher voltage distribution reserved for vessels with large propulsion or thruster loads, where LV cable sizes and switchgear would otherwise become unwieldy.

A cable is checked three separate ways, and whichever check demands the largest conductor is the one that decides the cable actually installed — checking only the first of the three is the most common error an inexperienced designer makes.

Cable tables quote ampacity at reference conditions — typically a single cable in free air at a moderate ambient. Real installations depart from that reference in two ways that both reduce how much current the cable can safely carry: a hotter ambient (an engine room, or a duct run near hot machinery) reduces the temperature margin to the insulation's limit, and running several loaded cables together in a bunch or tray reduces how effectively each one sheds heat to its surroundings. Both factors are applied as multiplying derating factors against the table figure, and both must be included — omitting the grouping factor because a run "looks like" a single cable in a duct full of others is a frequent oversight.

Volt drop is checked separately, using the actual load current over the real route length, and is commonly limited to a few percent (typically of the order of 6%) from the switchboard to the load. The concern is not the cable's safety but the equipment at the far end: a motor starting against a sagging supply develops less torque than its nameplate promises, and control gear, contactors and electronic equipment can misbehave or drop out below their rated voltage. A long, lightly loaded run can pass its current check comfortably and still fail on volt drop, particularly to consumers at the extremities of the ship.

Finally, the conductor must survive the fault energy it will see before its protective device clears — a short-circuit withstand check against the let-through energy of the upstream breaker or fuse, which is why protection settings and cable size are not independent decisions.

A ship's distribution system is protected in layers — final circuit breaker, sub-board incomer, main switchboard breaker, generator breaker — and discrimination is the requirement that a fault anywhere in that chain is cleared by the single device nearest to it, leaving everything upstream undisturbed. Lose discrimination and a fault on one feeder can trip several layers of breakers together, turning a localised problem into a much larger blackout than the fault itself ever justified.

Discrimination is achieved two ways together, not one: time grading and current grading. Time grading staggers the delay settings of successive devices moving upstream, so the nearest device has the shortest delay (or none at all — an instantaneous trip) and each device further back waits longer, giving the one closer to the fault the first chance to clear it. Current grading sets the pickup thresholds so that an upstream device's instantaneous element, if it has one, only operates for fault levels genuinely beyond what any downstream device is expected to clear — an instantaneous element on an intermediate breaker set below a fault current that a downstream breaker can also see will race that downstream breaker rather than back it up.

Because grading intervals accumulate moving back toward the generator, the plant's short-circuit withstand — how long the generator and its cabling can carry full fault current before damage — has to be checked against the worst-case total clearing time the grading scheme produces, not just against the fastest device's setting.

Most marine low-voltage distribution is deliberately insulated from the hull rather than solidly earthed, the reverse of common shore practice. In an insulated (IT) system, a single fault from one phase to the steel hull does not by itself create a short-circuit path, because the system has no intentional connection to earth for fault current to complete through; instead, an insulation-monitoring device detects the drop in insulation resistance and raises an alarm, and supply continues uninterrupted. That is the entire point of the arrangement at sea: a single earth fault, found on a vessel with no shore standby supply to switch to, must not itself cause a blackout.

The alarm is not the end of the matter, though — it is a warning that the system's protection against a second, more serious event has already been used up. If a second fault develops on a different phase before the first is traced and cleared, current can now flow phase-to-phase through the hull structure itself, and that is a genuine short circuit. Crews are trained, and watchkeeping procedures written, around finding and clearing the first fault promptly rather than treating the alarm as low priority because nothing has tripped.

Emergency power exists for the same underlying reason — normal generation can fail entirely, and essential services (steering, navigation lighting, fire pumps, communications, emergency lighting) cannot wait for it to be restored. An emergency generator, located and protected independently of the main machinery space, is arranged to start and pick up its load automatically within a short time of a blackout — commonly on the order of tens of seconds — and is sized to run the essential load list for a stated period of autonomy that scales with the type of ship and whether passengers are carried.

Shore connection is the other point where the ship's philosophy meets a different system: shore supply may differ in voltage, frequency or earthing arrangement from the ship's own, so a galvanic isolating transformer is normally fitted to keep the ship's insulated system genuinely isolated from shore earth, and phase sequence and synchronising checks are made before the ship's own generators are paralleled with, or handed over to, the shore feed.

The three examples below follow the chapter in order — sizing the generating plant from the load list, sizing a feeder cable against all three governing checks, and grading protection so a fault discriminates correctly — each with the full reasoning and every figure shown.