A chief engineer is judged less on running the plant day to day than on knowing why it was set up the way it was — and on reading the early signs that something in that setup has shifted.
Every main engine on a Class I ticket started life as a set of points on a load diagram. The engine builder does not offer one fixed rating — they offer a layout field, a parallelogram bounded by four corner points, L1 to L4, inside which a customer's chosen rated power and speed (the SMCR — specified maximum continuous rating) can sit. Move the rated point toward L1 and you get a smaller, faster-turning engine for the same power; move it toward L3 and you get a larger, slower-turning one that suits a bigger, more efficient propeller. The choice is made once, at contract stage, and everything else in this chapter follows from where inside that field the ship's engine was set.
Once the rated point is fixed, the propeller is matched to it so that the engine reaches rated power at rated rpm on a clean hull in calm water — in practice, slightly short of that, deliberately. Sea trials are always run "light": the propeller absorbs less power at any given rpm than it will once the hull has some months of fouling and the ship has met some weather. That gap is the light running margin, normally 4–7% in speed terms. It exists so that as the propeller curve gets heavier with time, the operating point drifts back toward — but does not cross — the torque/speed limit line running from L3 to L4. A ship delivered dead on its design curve, with no light running margin, would be running into that limit from the day it left the yard.
The light running margin is not slack in the engine — it is slack in the propeller curve. It is spent by fouling and weather, and a chief who does not track how much of it is left is the last to know the engine is about to run into its torque limit.
A shaftline is not rigid — it twists. Every power stroke in every cylinder puts a torque pulse into the crankshaft, and those pulses excite torsional vibration in the whole rotating system: crankshaft, flywheel, intermediate and tail shafts, and the propeller. Like any elastic system, the shaftline has natural torsional frequencies, and at certain engine speeds a firing-order harmonic coincides with one of them — a torsional critical. Where that coincidence falls inside the normal operating range, it cannot simply be ignored: the alternating stress it produces, sustained continuously, is a fatigue problem, not a noise problem.
The practical answer is a barred speed range (BSR): a band of rpm, established by torsional vibration calculation and confirmed by sea trial measurement, through which the engine is permitted to pass but never to run continuously or hover. Manoeuvring and controller logic are set up to accelerate or decelerate through the band quickly, and an engineer manually controlling speed near it needs to know where the band sits and treat it the same way. A torsional vibration damper, where fitted, is there to detune or absorb energy at the critical order and widen the safe range either side of the BSR — which is exactly why its condition (oil level and quality in a viscous damper, or wear in a spring or rubber element) is a survey and running-hours item, not a housekeeping one.
A barred speed range is a fatigue limit, stated as an rpm band because that is the only place it can be enforced from the bridge or the control room. Treat it as structurally as a maximum permissible torque, because that is what it is.
Deflection readings measure how much each crank web opens or closes as the crankshaft is rotated through a full revolution, using a dial gauge set between the webs. Taken at four or five crank positions per throw (commonly top, bottom, and two intermediate positions roughly 90° apart), they give a picture of how straight the crankshaft is sitting in its main bearings — which in turn reflects bedplate condition, chock tightness, main bearing wear, and how the hull itself is behaving under the engine.
The figures are compared against the builder's table of limits, throw by throw, and that comparison is necessary — but it is not sufficient. A single worn or low main bearing shows up as a localised change concentrated at that one throw and its immediate neighbours, with the rest of the engine essentially undisturbed. A smooth, progressive change that builds steadily from one end of the engine to the other — even while every individual reading stays inside the builder's limit — is a different finding altogether: it describes the crankshaft sagging or hogging as a whole, consistent with bedplate distortion, chock wear, changed tank contents, or the hull itself flexing under load. The figure that matters most is often not any single reading but the trend across adjacent throws.
Deflections are always read cold, with the engine stopped and, where practicable, with tank and mooring conditions similar to the baseline set, since loading and temperature both shift the readings independently of any real fault. A hot or running check, where fitted, adds a second, independent view of the same crankshaft under load.
One throw out of limit is a bearing question. A gradient across several throws, all still inside limit, is an alignment question — and it is the one that gets missed by an engineer working strictly throw by throw down the checklist.
Cylinder liners wear from two directions at once, and a chief who can tell which is dominant on a given engine is the one who can actually slow it down. Corrosive wear comes from sulphuric acid condensing on the liner wall: fuel sulphur burns to SO₂ and SO₃, which combine with water from combustion to form acid on the cylinder surface. The cylinder lubricant's base number (BN) exists specifically to neutralise that acid before it attacks the iron, so feed rate and BN both have to be matched to the fuel sulphur actually being burned, cylinder by cylinder condition and load. Too little oil, or oil with too low a BN for the fuel in use, leaves acid unneutralised and the liner etches; too much oil is its own problem, encouraging abrasive wear and driving up cylinder oil consumption and cost for no lubrication benefit.
Abrasive wear is mechanical: hard particles — principally aluminium and silicon oxide "cat fines" carried over from the refinery's catalytic cracking process — get between ring and liner and grind both away, along with piston ring grooves and ring butts. Cat fines are not removed by filtration in any meaningful sense at the concentrations found in residual fuel; they are removed by centrifugal purification, upstream of the engine, and that is the only place they can practically be dealt with.
Blaming worn liners on fuel sulphur, or on "bad fuel" generally, only accounts for half the mechanism. Feed rate, BN selection and purifier performance are the controllable half, and they are controlled onboard, not ashore.
Good combustion protects the liner as much as good lubrication does — a poorly tuned unit runs hotter, blows by more, and washes cylinder oil off the wall faster than a well-tuned one, independent of fuel quality. The standard onboard check is the spread between P_max (peak firing pressure) and P_comp (compression pressure alone, without injection) for each unit. A P_max − P_comp spread that is falling, or that differs noticeably from its sister units, points to late or poor injection, a leaking or worn injector, or a scavenge/exhaust problem on that unit specifically — it is a per-cylinder diagnostic, not an engine-average one, and it should be trended, not read once and filed.
Cat fines deserve separate attention because they are almost entirely preventable damage: aluminium and silicon particles from the refinery process that are abrasive at a hardness the liner and rings cannot resist. A bunker sample showing fines within the supply limit is not the same as fines reaching the engine within a safe limit — everything depends on what the purification system removes before the fuel gets there. A purifier has exactly three levers: throughput (a slower flow gives particles more residence time to be thrown out), operating temperature (set to the design point for the fuel's viscosity, not a fixed number), and settling or gravity-disc tank time upstream. Running a purifier fast to keep up with consumption, or cold to save steam, trades cat fine removal for convenience — and the bill arrives later, in liner and ring wear, not immediately.
Cat fines "as bunkered" and cat fines "at the engine inlet" are two different numbers. The purifier is what separates them, and its settings are a daily operational decision, not a fixed piece of equipment simply doing its job.
Where a ship cannot meet its attained EEXI through hull or hydrodynamic measures alone, the common remaining route is to limit the power the engine is permitted to deliver — either engine power limitation (EPL), which restricts the fuel rack or index so the engine cannot develop more than a set percentage of its original SMCR at any rpm, or shaft power limitation (ShaPoLi), which limits shaft power directly, typically through the governor and a torque/power limiter, and can be more flexible about which combinations of torque and rpm are restricted. Both reduce the maximum power on the specification the ship is certified against, and both are sealed: the limitation is not meant to be something a crew can quietly wind back up when it is inconvenient.
Both also carry a safety override, because a class society and flag state accept that there are situations — heavy weather, restricted waters, an emergency — where the extra power exists on the engine and withholding it would be the greater risk. The override is not a loophole; it is a deliberately built-in safety valve, on the explicit condition that its use is logged and reported. Used and recorded, it is exactly what it was designed for. Used and not recorded, the ship's compliance basis is compromised regardless of whether the engine was ever actually at risk mechanically.
For the chief, day to day, the practical effect of either scheme is a smaller margin above the normal service point for heavy weather than an unrestricted engine of the same size would have. Knowing which scheme is fitted, where the sealed limit sits, and how much reserve the override actually restores is part of knowing the plant — not paperwork bolted on afterward.
A power limitation scheme does not just cap a number — it resizes the safety margin the ship has left for weather. An override used without being logged fixes the immediate problem and creates a compliance one.
The three problems below step through the reasoning a Class I candidate is expected to show: converting a stated margin into the power it actually buys, reading a deflection trend rather than a single figure, and carrying an EEXI power limitation through to the rpm and reserve it leaves for heavy weather.
A vessel's engine has SMCR = 13,560 kW at 105 rpm, the rated point sitting inside the layout field. On sea trials, light running, the engine reaches its rated power of 13,560 kW at 111.3 rpm. Years later, fouling and average weather mean the actual propeller curve now demands 15% more power than the light (trial) curve did, at the same 105 rpm. Find how much power margin the engine has left at 105 rpm before it reaches MCR.
SMCR = 13,560 kW at 105 rpm (rated point, layout field) Light running: MCR power (13,560 kW) reached at 111.3 rpm on trial In service: propeller now needs 15% more power at a given rpm than the light (trial) curve Propeller law: P ∝ N³ along a given curve
Find how much power margin the engine has left at 105 rpm before it reaches MCR
Step 1 — read off the light running speed.
A light running margin is expressed in rpm: on trials, clean hull, no weather, the engine reaches rated power at a higher speed than the design rpm. 111.3 rpm is that trial speed.
Step 2 — carry that same power back down to the rated speed.
On the same light curve. The trial curve obeys the cube law, so the power it would absorb at 105 rpm — short of where it actually reached MCR — follows from the ratio of speeds cubed.
Step 3 — express that as a power margin.
The gap between MCR and this figure is what the 6% speed margin is actually worth at the rated point.
Step 4 — apply what fouling and weather have since cost.
The real propeller curve now demands 15% more power than the light curve did, for the same 105 rpm.
Step 5 — compare with MCR to find what is left.
AnswerThe engine can still reach 105 rpm continuously — about 467 kW (3.4% of MCR) of margin remains before the torque/speed limit line is reached — but the original 6% speed margin, worth 16% of MCR when new, has been almost entirely used up. Hull cleaning or docking should be scheduled before the next heavy-weather season, not after the margin runs out.
The trap: treating a light running margin as a fixed rpm buffer instead of converting it through the cube law — a 6% speed margin is not a 6% power margin, and confusing the two over- or under-states how much reserve is genuinely left.
A six-cylinder low-speed engine's vertical web deflections (top gap minus bottom gap, in millimetres) were logged at commissioning and repeated at the current survey. The builder's limit for this engine is ±0.35 mm per throw. Assess whether the current readings indicate a problem, and if so, what kind.
Vertical deflection at commissioning (mm): #1 +0.02, #2 +0.01, #3 0.00, #4 −0.01, #5 −0.02, #6 −0.03 Vertical deflection now (mm): #1 −0.01, #2 −0.03, #3 −0.05, #4 −0.08, #5 −0.11, #6 −0.14 Builder's limit: ±0.35 mm per throw
Step 1 — check each throw against the builder's limit.
| Throw | Commissioning | Now | Within ±0.35 mm? |
|---|---|---|---|
| 1 | +0.02 | −0.01 | Yes |
| 2 | +0.01 | −0.03 | Yes |
| 3 | 0.00 | −0.05 | Yes |
| 4 | −0.01 | −0.08 | Yes |
| 5 | −0.02 | −0.11 | Yes |
| 6 | −0.03 | −0.14 | Yes |
Taken one at a time, none of the six current readings looks alarming.
Step 2 — find the change at each throw since commissioning.
Not just today's value.
Step 3 — look at the shape of the change across the engine.
Not each figure in isolation. A single worn bottom-end bearing shows up as a step change at one throw that does not carry through to its neighbours. Here the change grows smoothly from throw 1 to throw 6.
Step 4 — interpret the pattern.
A steady, cumulative drift along the length of the engine, with every reading still inside the builder's limit, is the signature of a whole-engine bending trend — bedplate or chock condition, tank loading, or hull working — not a single degraded bearing.
AnswerNo individual throw exceeds ±0.35 mm, so the readings would pass a throw-by-throw check — but the smooth, cumulative drift (about 0.11 mm of extra sag by throw 6, growing at roughly 0.016 mm per throw) points to a developing hull/bedplate alignment issue. Recommend a hot (running) check, a review of chock tightness and tank/ballast condition, and closer monitoring at the next opportunity, rather than opening up a bearing.
The trap: ticking off each of the six figures against the builder's limit and stopping there — every single reading can be compliant while the trend across the engine is already telling you something is moving.
A vessel's engine has an original (unlimited) SMCR of 13,560 kW at 105 rpm, on a propeller curve that follows the cube law. To bring the ship's attained EEXI within the required value, the owner applies engine power limitation (EPL), sealing the fuel rack so the engine cannot exceed 85% of the original SMCR. Normal continuous service rating (CSR) is set at 90% of this new, limited MCR. In heavy weather the chief authorises use of the sealed override to recover full original power for a period, to keep adequate way on. Find the limited MCR, the CSR, and the size of the override — and state what the engineer must do as a result.
Original SMCR = 13,560 kW at 105 rpm Propeller curve: P ∝ N³ EPL sets limited MCR = 85% of original SMCR CSR = 90% of the limited MCR Override recovers 100% of the original SMCR for a period
Find the limited MCR, the CSR, and the size of the override — and state what the engineer must do as a result
Step 1 — find the limited MCR.
EPL caps what the fuel rack can deliver; it does not change the hull or the propeller, so the engine still runs down the same propeller curve, only to a lower ceiling.
Step 2 — find the rpm that goes with it.
Because the ship still follows the same P ∝ N³ curve, the limited power point corresponds to a lower rpm, not the original 105 rpm at reduced fuel.
Step 3 — find the CSR.
The day-to-day service point carries its own margin below the new, limited MCR, exactly as CSR is normally set below an unrestricted MCR.
Step 4 — quantify the override.
Recovering the full original SMCR means running above the sealed limit, not above the engine's mechanical capability — the engine was always capable of 13,560 kW; only the fuel rack setting changed.
AnswerThe limited MCR is 11,526 kW at about 99.5 rpm, with CSR at about 10,373 kW / 96.0 rpm. Using the override brings the engine back to its full 13,560 kW — about 17.6% above the sealed limit. The engine is mechanically able to do this; what makes it acceptable is that the override, its reason and its duration are logged and reported, restoring the vessel's compliance record. Used without that record, it is a compliance failure even though nothing was ever overloaded.
The trap: assuming that because the engine can safely deliver full power again, using the sealed override without logging and reporting it is a technicality — the failure here is administrative, not mechanical, and it invalidates the power-limitation basis the ship is certified against.
P ∝ N³ along a propeller curveLight running margin (typically 4–7% in speed) converts to a larger % of powerLayout field: L1–L4SMCR chosen inside it; the L3–L4 line is the torque/speed limit, separate from L1–L2Barred speed range (BSR)Torsional critical inside the operating range — pass through, never dwell or hoverTorsional vibration damperDetunes/absorbs the critical order; its condition is a survey and running-hours itemDeflection readings4–5 positions per throw; compare against the builder's table and the trend across adjacent throwsCylinder oil feed = g/kWh, matched to BNBN neutralises acid from fuel sulphur; feed rate and BN both matterCat fines (Al+Si) ≤ 15 mg/kg as bunkeredA supply limit, not an at-engine guarantee — purifier removes the restPurifier leversThroughput, temperature, settling/gravity-disc time — the only three controls on cat fine removalP_max − P_comp spreadPer-cylinder trend for injection timing/condition, not an engine averageEPL / ShaPoLi + overrideEEXI power/shaft limitation route; safety override permitted but must be logged and reported