Light-load diesel running

Why does a ship's engineer worry more about an engine run gently for weeks than one worked hard?

Everything we believe about machinery says gentle use is kind use: half the throttle, half the wear. Yet an engineer watches an engine ambling at low load through weeks of slow steaming with more anxiety than one held near its rating, and eventually goes looking for an excuse to load it up hard.

If wear rose with load, weeks at low load ought to be the cheapest weeks the engine ever has. So what is the engine doing at high load that it stops doing at low load — something whose absence is the damage?

Ask what load supplies inside a cylinder. Not just work at the shaft: peak pressure and temperature, both of which fall steeply when fuelling falls. Two quite different things depend on them.

The first is the piston ring seal, which is not the simple spring most people imagine. A ring's outward force against the liner is largely supplied by combustion gas passing behind it and pushing it out — it is pressure-energised, so its sealing force is roughly proportional to the very pressure it seals against. At low load there is little pressure to energise it, so the seal is slack, more gas blows past, and the ring's rubbing action on the liner is weak.

The second is temperature. Combustion at low load is cooler and less complete, and the liner is cooler too, the cooling system being sized for full power and indifferent to your asking for a quarter of it. Cool surfaces plus incompletely burnt heavy fuel means unburnt hydrocarbons and soot reach the cylinder wall instead of being consumed.

Now put those together and something specific happens to a surface that was deliberately made rough. A cylinder liner is honed with a crosshatch of fine grooves whose job is to hold oil and let the rings bed, and deposits settling into them are ordinarily scoured out by the ring's own pressure-energised rubbing. Below some combination of temperature and pressure, deposition outruns scouring; the grooves fill with a hard baked varnish and the liner goes smooth and shiny. That is bore glazing.

Why is a shiny bore bad? Because the grooves were the oil control. A glazed liner neither holds an oil film where it is wanted nor lets the rings seal, so oil passes up into combustion and combustion gas passes down into the crankcase — and more blowby means more contamination and worse sealing, which means more deposit. It feeds itself, which is why engineers speak of a point of no return.

The rest of the low-load complaints are the same incomplete combustion showing up elsewhere: a turbocharger spun slowly by weak exhaust energy gives less boost and so less air, making combustion worse still; unburnt fuel dilutes the oil it drains past; and soot accumulating in uptakes and, in large two-strokes, in the scavenge spaces is a genuine fire hazard rather than a housekeeping nuisance. On generator sets the same picture has its own name — wet stacking.

I will not quote a minimum load figure, because there is not a single honest one: builders publish a minimum sustained load and a maximum period below it, and the numbers differ by engine, fuel and installation. The shape of the rule is what matters — there is a threshold, and time below it accumulates.

Two refinements keep this honest.

The first is that this is not the same story as running an engine in, though the physics is a cousin. Run-in is two surfaces becoming the shape of each other, and moderate load helps because ring pressure does the bedding. Light-load damage is the same lever pulled the other way: the geometry running-in produced gets filled and polished away because the pressure that maintains it is absent. One is convergence, the other is loss of what converged.

The second is the threshold and its one-way gate. Risk is not proportional to how gently you run; below a certain cylinder condition the sign of the process changes, from self-cleaning to self-fouling. And the recoverable window is limited: soft deposit can still be burnt and scoured off by a deliberate period at high load, which is why engineers arrange load-ups and why generator sets are exercised against load banks. Once the deposit has hardened and the crosshatch is gone, loading up no longer helps — the surface that would have done the scouring is the surface that was lost. Some of the older folklore does deserve scepticism, though: modern engines with common-rail injection, variable-geometry turbochargers and better detergent oils tolerate low load considerably better than the plant this rule of thumb grew up around.

Here is the test that would break the account. If the driver is cylinder temperature rather than load as such, measures that raise cylinder temperature without raising load should prevent the damage: running the jacket water hotter, preheating charge air, or cutting out cylinders so the rest carry proper load. Those are precisely the mitigations in use, which supports it. The refutation would be an engine held at correct liner temperature and air-fuel ratio at low load that glazed at the same rate as a cold one — that would move the explanation onto ring contact force alone and make the thermal story decoration.

• Why slow steaming forced builders to re-engineer cylinder lubrication and turbocharging rather than simply warn against low load.
• How scavenge-space and uptake fires start, and why they are the acute form of the same deposit problem.

Every term the collection defines is gathered in the glossary.