Auxiliary Machinery & Systems

Ask why steering gear draws so much attention in an MEO Class I oral and the answer is always the same: it is the one system on board where a failure is visible to the bridge, the pilot and anyone watching from the shore, within seconds. The rules are written around that visibility rather than around elegance, and the examiner wants to know you understand why the numbers are what they are, not just that you can recite them.

The headline figure is meaningless without its conditions: maximum ahead service speed and the ship's deepest seagoing draught, together, because both push the rudder into the worst hydrodynamic loading it will ever see. A test performed alongside the berth, slow and light, proves almost nothing, since the torque the gear has to overcome rises with the square of the speed past the rudder. Where two independent power units are fitted, each one has to be capable of meeting the main timing on its own — not the two working together — and that is the detail candidates forget: it is easy to test with both pumps running and never ask what happens with only one.

• Two power units — each must independently meet the 28-second timing, not only the two combined.
• Emergency power supply — picks the steering load up quickly if the main supply is lost.
• Rudder angle indication — taken directly from the rudder stock so it survives the same failure that might take out a remote system.
• Testing regime — full test within 12 hours of departure, plus an emergency steering drill at least every three months.

Cavitation is not really a pump problem — it is a pressure problem that happens to occur inside one. At the eye of the impeller the fluid is accelerated and the local static pressure drops; if that pressure falls to the fluid's own vapour pressure at the prevailing temperature, the liquid boils, not from heat but from a fall in pressure, and tiny vapour bubbles form. Those bubbles are swept into the higher-pressure region further round the impeller, where they collapse violently. The collapse is what pits vanes and casings and produces the characteristic gravel-like noise; the vapour formation upstream of that is the actual cause.

NPSH available is what the suction side of the system actually offers — atmospheric or tank pressure converted to a head, minus the liquid's vapour pressure head at running temperature, plus or minus the static height between liquid surface and pump, minus suction-line friction. NPSH required is a property of the pump itself, stated by the manufacturer from test, and it rises with pump speed and flow. The rule is simple to state and easy to forget under exam pressure: available must exceed required, with a margin, because neither figure is worth running to zero against.

Hot condensate and light distillates catch pumps out for the same reason in both cases: their vapour pressure at running temperature is already close to atmospheric, so it takes very little extra pressure drop at the impeller eye to reach the local boiling point. A pump that loses suction shortly after a fuel changeover, or a condensate pump that starts vibrating as the plant warms through, is very often behaving exactly as the NPSH margin predicts, not failing mechanically.

Starting air capacity is sized by how many consecutive starts the receivers must deliver without the compressor topping them up, because a ship that can start once and then has to wait for recharging is not a ship you can manoeuvre with confidence in a tight channel. The convention most candidates learn is twelve starts for a reversible, direct-drive engine and six for a non-reversible one — the reversible engine needs more because reversing itself consumes air on top of simply getting the engine running, and a failed reversal close to another vessel is exactly the moment you cannot afford to be short. Capacity is normally split between two receivers of broadly equal size rather than one large one, so a single receiver out of service for survey still leaves the engine able to start.

The protection fitted to the starting air line exists because of one specific failure chain, and it rewards understanding the sequence rather than reciting a parts list. A starting valve that is leaking — worn, or stuck slightly open — allows hot combustion gas to blow back into the starting air manifold instead of the manifold only ever pushing air into the cylinder. That hot gas can ignite oil residue or carbon deposits inside the pipework, and because the pipe is full of compressed air an explosion can propagate along the line toward the receiver.

• Non-return valve — fitted close to each cylinder's starting valve, stopping combustion gas reaching the common manifold in the first place.
• Bursting disc — a deliberately weak point that relieves an explosion before it can rupture the pipe elsewhere.
• Flame arrester / relief device — checks the flame front travelling further along the manifold.

A fresh water generator turns seawater into fresh water using heat that is, by power-plant standards, nearly free — jacket cooling water coming off the main engine, rather than raising fresh steam specifically for the purpose. The only way to boil seawater with heat that cool is to drop the pressure it boils under, and that is exactly what the evaporator shell does: an ejector or vacuum pump holds the shell well below atmospheric pressure, lowering the boiling point of the seawater inside to somewhere the available heat source can sustain. Reduce the pressure far enough and heat that would otherwise be rejected overboard becomes the entire energy source for the plant.

Operating under vacuum brings a second benefit alongside the cheap heat: many scale-forming compounds in seawater, calcium sulphate in particular, become less soluble as temperature rises, so a plant boiling at a lower temperature under vacuum scales more slowly than one running hot. That matters because scale on the heating surfaces is the main enemy of output — a scaled surface transfers heat poorly, and the crew is left wondering why a plant that made twenty tonnes a day last month is struggling to make twelve.

Throughput and quality trade against each other here just as they do in a purifier. Push the plant harder than its design output and priming becomes a real risk — droplets of concentrated brine carried over into the distillate rather than left behind in the shell — so product salinity is monitored continuously, with an automatic dump valve diverting any batch that reads high away from the fresh water tank. A generator making water fast and passing every reading is doing its job; one making water fast and occasionally spiking on salinity is one setting away from contaminating the tank it feeds.

A centrifugal purifier separates water and solids from fuel or lubricating oil by spinning the mixture hard enough that the density difference between oil, water and solids does the sorting job gravity alone would do far too slowly. Heavier water and solids move outward to the bowl wall while the lighter clean oil moves toward the centre and out through the discharge — and how well that works comes down to how long a given parcel of liquid spends inside the bowl under that force, its residence time.

Push more throughput through the same bowl and residence time falls, because the same separating space now has less time to act on more flow. Water droplets and cat fines that would have migrated to the bowl wall in time are instead carried out with the clean oil. This is why raising throughput to keep pace with higher fuel consumption is a trap rather than a solution: it appears to solve a supply problem while quietly degrading separation at the moment the engine most needs clean fuel. Cat fines — abrasive aluminium and silicon particles left from the catalytic cracking process — are dense enough to separate readily given time, but a purifier run too fast passes a meaningful fraction straight through, and the resulting liner and ring wear looks, on a wear-rate trend, exactly like a fuel quality problem rather than a purifier setting error.

Heating the feed works in the purifier's favour because it lowers viscosity, letting the density difference act faster — why fuel and lube oil purifiers run hot rather than cold. The gravity disc completes the picture: its size sets where the oil/water interface sits inside the bowl, chosen for the density of the oil being processed. Fit a disc sized for the wrong density and the interface moves to the wrong radius, either letting water through with the oil or losing oil out with the water and sludge.

Most machinery on a modern ship is cooled by fresh water, not seawater directly, and the reason is corrosion management rather than performance. Seawater fouls, scales and corrodes far faster than treated fresh water, so the design confines it to as few components as practically possible — typically one or two central coolers — and lets a closed fresh water circuit carry the heat on to the engine jackets, charge air coolers, lubricating oil coolers and the rest of the plant. Losing a seawater pump or fouling a seawater strainer then threatens one cooler rather than every piece of cooled machinery on board, and the fresh water side can be treated chemically in a way that would be uneconomic to do to the sea.

Hydraulic systems do a different kind of work — steering gear, cargo gear, hatch covers, remote valve operation and controllable-pitch propeller control all rely on a closed circuit of oil under pressure rather than a rotating shaft or a cable run, because hydraulics deliver large forces through pipework routed almost anywhere in the ship. The trade-off is that the whole system depends on cleanliness and on the integrity of that pressure boundary: a contaminated filter or a failed seal does not just reduce performance gradually, it can take the function out of service quite suddenly — exactly why a system like steering gear is never left running on a single hydraulic power unit.

Every example in this chapter answers the same underlying question from a different angle: what still works when one thing has already failed? A standby pump that starts automatically on falling pressure, a cross-connection that lets one cooling or fuel system back up another, an emergency fire pump sited outside the machinery space with its own sea suction and its own independent power supply — none of these exist because the primary equipment is expected to fail often. They exist so that when it does fail, the consequence is contained rather than total. Read an unfamiliar auxiliary system question from that direction — what is the backup, and what specifically does it not share with the thing it is backing up — and most of the answer falls into place before you reach for a formula.

The three problems below step through the numbers most MEO Class I candidates only meet as bare formulas — NPSH margin on a hot pump duty, starting air redundancy when a receiver is out of service, and steering gear timing with one power unit rather than two. Work through the given data before reading the steps.

Worked example 1

NPSH available vs required on a hot condensate pump

A condensate pump draws from a hotwell under a flooded suction. The chief engineer wants to know, before the plant is brought up to full load, whether this duty is going to cavitate — and if it will, how much extra static head would put a comfortable margin back into the calculation.

Given

Barometric pressure equivalent head, p_a/ρg = 10.3 m Vapour pressure of the condensate at running temperature, p_v/ρg = 9.7 m Hotwell water level above pump centreline (flooded suction), h_s = +1.2 m Friction loss in the suction line, h_f = 0.4 m NPSH required by the pump at this duty, NPSH_r = 1.6 m Desired margin above NPSH_r = 0.5 m

1. Work out NPSH available from the heads given.

   NPSH_a=(p_a − p_v)/ρg + h_s − h_f =(10.3 − 9.7) + 1.2 − 0.4 =0.6 + 1.2 − 0.4 =1.4 m

   Atmospheric head less vapour pressure head, adjusted for the static height of liquid above the pump and the friction lost reaching it.

2. Compare against NPSH required.

   Margin=NPSH_a − NPSH_r =1.4 − 1.6 =−0.2 m (deficit)

   The pump needs 1.6 m and the system is only offering 1.4 m, so there is a deficit before any margin is even considered — this duty will cavitate, not just run tight.

3. Find how much extra static head would restore a 0.5 m margin above NPSH required.

   Required NPSH_a=NPSH_r + margin =1.6 + 0.5 =2.1 m Extra head needed=2.1 − 1.4 =0.7 m

   For example by raising the hotwell mounting relative to the pump.

AnswerThe pump is 0.2 m short of NPSH required and will cavitate at this duty; raising the hotwell level (or otherwise cutting suction losses) by about 0.7 m relative to the pump would restore a 0.5 m margin.

The trap: treating 1.4 m available against a 1.6 m requirement as "close enough". NPSH margin is available minus required — a negative result means cavitation at this duty, not a tight but acceptable one.

Worked example 2

Starting air receiver capacity and single-receiver redundancy

A non-reversible, direct-drive main engine is served by two identical starting air receivers. One receiver is isolated for internal survey while the ship is alongside. The second engineer wants to know whether the engine can still be started legally on the remaining receiver alone, and how much spare capacity the installation actually carries above the bare rule.

Given

Each receiver volume, V_r = 4 m³ (two receivers fitted) Charged pressure, P1 = 30 bar (absolute) Minimum working pressure for reliable starting, P2 = 6 bar (absolute) Free-air consumption per start, Q_start = 8 m³ (at 1 bar) Engine: non-reversible, direct-drive

1. Find the free-air volume one receiver can deliver as it blows down from charge pressure to the minimum working pressure, taking the expansion as isothermal so free-air volume scales directly with the pressure drop.

   V_free=V_r × (P1 − P2) =4 × (30 − 6) =4 × 24 =96 m³ (free air)

2. Convert to starts using the free-air consumption per start.

   Starts (one receiver)=V_free / Q_start =96 / 8 =12 starts

   For one receiver working alone — the case that matters with the other isolated for survey.

3. Check that against the regulatory minimum for a non-reversible engine.

   12 starts>6 required ⇒ compliant on one receiver alone Minimum receiver volume for 6 starts: V_r(min)=(6 × Q_start) / (P1 − P2) =(6 × 8) / 24 =48 / 24 =2 m³

   Then work backwards to the smallest receiver that would exactly meet that minimum under the same pressures, to see how much margin the actual installation carries.

AnswerOne receiver alone delivers 12 starts — twice the six-start minimum for a non-reversible engine — so the plant stays fully compliant with either receiver isolated. The fitted 4 m³ receiver is twice the 2 m³ that would bare-minimum satisfy the rule; the rest is margin, not oversizing.

The trap: adding both receivers' capacity together and checking that combined figure against the six-start minimum. The redundancy requirement is that each receiver independently covers the minimum, because a receiver under survey is not available to add to the other's total.

Worked example 3

Steering gear timing with one power unit versus two

A hydraulic steering gear is being assessed at the design stage. The oil volume needed to swing the rudder through the full 35°P to 30°S movement, and the rated flow of each of its two identical power units, are known. Find the time to complete the swing with both units running, and with only one, and judge each against the regulatory 28-second limit.

Given

Oil volume to swing rudder 35°P to 30°S (65° of travel), V = 1.68 m³ Rated flow of each power unit at working pressure, q = 0.06 m³/s Two power units fitted Regulatory limit = 28 s, at max ahead service speed and deepest seagoing draught

Required

Find the time to complete the swing with both units running, and with only one, and judge each against the regulatory 28-second limit

1. Find the time to complete the swing with both power units running together.

   t (two units)=V / (2q) =1.68 / (2 × 0.06) =1.68 / 0.12 =14 s

   Since their flows add directly into the same ram.

2. Now repeat with only one power unit running.

   t (one unit)=V / q =1.68 / 0.06 =28 s

   The case a chief has to check even though it is rarely the one demonstrated at the dock trial, because it is the case the redundancy requirement exists for.

3. Compare both results against the 28-second limit and draw the design conclusion.

   Margin (two units)=28 − 14 = 14 s Margin (one unit)=28 − 28 = 0 s

AnswerWith both power units running the gear meets the requirement with a comfortable 14-second margin; with only one unit in service it meets the 28-second limit exactly, with zero margin. That zero-margin case is precisely why two independent power units, each able to meet the timing alone, are required rather than two units sized only to meet it together.

The trap: demonstrating the 28-second requirement only with both power units running — the easy case — and never checking the single-unit condition that the redundancy requirement actually exists to cover.