Propulsion & Power Generation

Choosing a prime mover starts with the ship's actual speed-time profile, not with a single design point on a curve. A product tanker running flat out for weeks between coasts, a supply vessel spending half its life on dynamic positioning, and a ferry doing forty short hops a day are three different sizing problems even if their installed power comes out similar on paper.

Slow-speed two-stroke diesels, direct-coupled to the propeller with no gearbox, remain the default for large deep-sea ships: they turn at propeller-friendly revolutions, burn heavy fuel, and post the best thermal efficiency of any marine prime mover across most of their range. The price is a single point of failure — lose that one engine and there is no propulsion — and a physically long, heavy installation.

Medium-speed four-stroke diesels turn faster than any propeller wants to, so they need a reduction gearbox, but they are lighter and more compact for a given power, and it is easy to fit two or more of them on one gearbox for redundancy or to shut one down and run efficiently at part load. Gas turbines go further in the same direction: very high power density and fast starts, at the cost of poor part-load specific fuel consumption and a real appetite for clean fuel, so they tend to appear either alone on fast, light craft or alongside diesels in combined plants that hand off between the two as speed demands.

A fixed-pitch propeller absorbs power in a very particular way. Its thrust and torque coefficients depend on the advance coefficient — essentially the ratio of ship speed to propeller rpm — and for a ship running close to its normal speed-length relationship, that ratio stays roughly constant as revolutions change. Work the non-dimensional groups back out and, for one particular hull and fouling condition, torque comes out proportional to the square of revolutions and power to the cube.

The catch is that word "one." A clean hull on trial and a fouled hull two years into a docking cycle are different curves entirely, both obeying the cube law individually but sitting at different heights. On the fouled, heavier curve, a given rpm demands more torque than it did on the clean one. Engines, meanwhile, have their own limit: a maximum torque they can deliver, set by the turbocharger and fuel system, which is reached before the rpm limit if the propeller curve is heavy enough.

The light running margin is the gap built in to cover that difference. The engine and propeller are matched so that MCR sits on a design ("heavy") curve that already assumes some in-service fouling; on trial, with a clean hull, the same engine reaches MCR power at a noticeably higher rpm — typically 4–7 % above rated. That headroom is what keeps the operating point clear of the torque limit as the hull fouls, the draught changes between loaded and ballast condition, or the ship pushes into heavy weather, all of which move the propeller curve toward "heavier."

The shaft line carries three distinct loads at once, and a diameter chosen for one of them can still be wrong for another. Torsion, from the torque the engine delivers, sets the basic diameter through a straightforward strength check. Thrust, the axial push the propeller generates, has to be carried back through a thrust bearing to the hull structure rather than through every intermediate bearing along the way. And the whole line, from crankshaft to propeller, behaves as a torsional spring-mass system with its own natural frequencies.

That formula also carries a useful piece of intuition: diameter only grows with the cube root of torque. A modest increase in rated power costs relatively little in shaft diameter; a large one costs a great deal, because the relationship is anything but linear once you work back from stress to size.

Alignment matters as much as diameter. A shaft line running through several bearings — main engine, gearbox, one or more intermediate bearings, stern tube — has to share its weight and thrust sensibly across them; get the alignment wrong and one bearing carries far more than its share while another barely touches its shaft, showing up later as abnormal wear or a hot bearing rather than as an obvious fault at commissioning.

Torsional vibration is the least intuitive of the three. The engine's firing impulses excite the shaft line's torsional natural frequencies, and where one of those frequencies falls inside the normal operating speed range, running continuously at that rpm can drive vibratory stress well beyond what steady torque alone would suggest. The fix is not to redesign the shaft for every ship — it is to identify the offending rpm band and bar the engine from dwelling in it, passing through quickly on the way to a different speed instead.

Propeller efficiency generally improves with a larger, slower-turning propeller, up to the point where blade tip speeds start inviting cavitation. Medium- and high-speed engines turn far faster than that optimum, so a reduction gearbox sits between engine and shaft to let each side run at the speed that suits it — the engine near its rated rpm, the propeller at the rpm the powering estimate actually calls for.

A single engine driving through a single reduction stage is the simplest arrangement. Multiple-input gearboxes, combining two or more engines onto one output shaft, add flexibility: run both engines for full power and speed, or declutch one and run on the other alone for slow-speed economy or maintenance, without losing propulsion altogether. That same declutching ability is what lets a combined diesel-and-gas-turbine plant hand off between prime movers as the speed demand changes.

Flexible couplings between engine and gearbox absorb the torsional irregularity that comes from individual cylinder firing pulses and take up the small misalignments that inevitably exist between two separately mounted machines. Gear teeth themselves are sized and lubricated to a defined duty cycle, and because a gearbox is comparatively unforgiving of misalignment, the whole line's alignment tolerance often ends up governed by the gearbox rather than by the shaft bearings either side of it.

Diesel-electric propulsion breaks the direct mechanical link between prime mover and propeller entirely. Engines drive generators; generators feed a common electrical bus; propulsion motors, fed from that bus, turn the propeller at whatever speed the ship needs. Engine speed and propeller speed are no longer the same question.

That decoupling buys three things at once. Each generating set can run near its own best-efficiency point regardless of ship speed, rather than being dragged up and down a single engine's efficiency curve as the propeller demands vary. Load can be shared across several smaller sets instead of carried by one large engine, so at reduced ship speed some sets can be shut down altogether rather than all of them idling at low, inefficient load. And because propulsion no longer depends on one specific engine surviving, losing a generating set is a capacity reduction rather than a total loss of propulsion — valuable on vessels that need to hold position or manoeuvre precisely, such as those with dynamic positioning.

None of that comes free. Every conversion — mechanical to electrical at the generator, electrical to mechanical again at the motor, plus whatever power electronics sit in between — sheds a little energy as heat, and those losses stack. The electrical plant itself becomes larger and more complex: more switchboards, more protection, more places for a fault to hide. Hybrid arrangements try to take the useful parts of this without going fully electric — a shaft generator or shaft motor bolted onto an otherwise conventional shaft line, or a battery bank buffering transient loads so the engines can be run at steadier, more efficient output and, in port, for brief periods of quiet, zero-emission running.

Sizing the ship's generators starts with a table, not a rule of thumb. Every electrical consumer is listed with its rated power and a load factor for each operating condition — at sea, in port, manoeuvring, emergency — reflecting how hard it actually runs rather than its nameplate rating. A steering gear motor rated at a hundred kilowatts might run at a fraction of that most of the time at sea and briefly much harder while manoeuvring; a continuous cooling water pump runs close to its rating whenever it runs at all. Multiplying rating by load factor and summing the column gives the absorbed load for that condition — the figure that actually matters.

Generators are then sized on whichever condition turns out to be the worst — often manoeuvring or an emergency scenario rather than the steady at-sea condition — with a design margin added for growth and estimating uncertainty, and enough sets specified so that one can be down for maintenance or repair without shedding load. Sizing on the sum of connected nameplate ratings instead of the load-balance figure is a common shortcut, and it is a poor one: it ignores diversity between consumers and tends to land on a plant that is bigger, and more often lightly loaded, than the ship actually needs.

Fuel consumption follows once the power is known, using the specific fuel oil consumption quoted for that engine at that load. The figure only means anything, though, once it has been corrected to a common reference — ambient temperature, pressure, humidity and fuel calorific value all shift a test-bed SFOC number, and comparing two engines' consumption figures without correcting both to the same reference conditions is comparing two different tests, not two different engines.

The three examples below work through the mechanisms above with real numbers: a light running margin that has quietly turned negative, a shaft diameter checked against both its current duty and a proposed re-rating, and a full load balance used to size the sea-going generating plant.