The Cheapest Lever on the Ship, and What Its Percentage Is Of

Every fleet circular that has ever carried the words fuel efficiency carries this sentence somewhere: trim optimisation saves two to four per cent, it costs nothing, do it. That range is what a voyage-optimisation supplier quotes when asked what its trim module contributes, which makes it vendor-flavoured, and there is no document behind it to open.

Then, in the same week, somebody sends you a study reporting fifteen per cent, a class-society sea trial reporting nine, and an IMO page reporting half of one per cent. Thirty to one, between figures that are all published and all honest. Nobody on that list is lying: they are answering different questions, and telling those questions apart is this lesson.

The staircase, in descending order

FORCE Technology has run trim tests on almost 300 vessels: tankers, container vessels, LNG carriers, ro-ro vessels and ferries. Reichel, Minchev and Larsen report the findings in Trim Optimisation - Theory and Practice (TransNav 8(3), 2014, free PDF), and they put both numbers in one paragraph, three lines apart: testing so far shows "possible fuel savings of up to 15% at specific conditions compared to even keel", and then, "In overall fleet operations, typical savings can be as high as 2 to 3%". One firm, one body of data, two figures five times apart on one page.

The China Classification Society took a 4,250 TEU container ship in service from Nansha to Tianjin. She sailed at 8.7 metres draught; ballast was increased by 1,800 tonnes to bring her 0.8 metres by the bow; and at unchanged speed "the required main engine power is reduced by 910KW", with 14 tonnes of fuel saved on the passage and an energy saving rate of 9.2% (CCS, open). Notice the shape of the claim before the size of it: one ship, one loading condition, one passage, and a power reduction at constant speed rather than a fuel figure over a year. CCS then step down a rung themselves, putting her annual potential at 5 to 8 per cent, averaging 3.2 tonnes a day.

IMO's own GreenVoyage2050 trim and draft optimisation page (open) is the most conservative document in the industry on this subject, and it quietly does the job this lesson is about. "Optimized trim and draft are estimated to be able to reduce fuel consumption on the main engine by 0.5 - 2% for the vast majority of ship types." Then a carve-out: "For ships that often use partial loads (e.g. container, ro-ro, etc.), the effect can be up to 5% in sailing mode." Then a measurement: "For PSVs, a saving of 4% has been verified, comparing two sister ships operating with and without optimal trim."

Per cent of what

None of those three documents disagrees with any other. Laid side by side the pattern is immediate: they are percentages of different things, and the industry uses one word for all of them.

There are three denominators in circulation. One condition: FORCE's fifteen and CCS's 9.2 are both the gap between one trim and another, at one draught, one speed, one afternoon. One voyage or one mode: GreenVoyage2050's "up to 5% in sailing mode" and its verified 4 per cent on a pair of PSVs are averages over a way of operating, wider than a condition and narrower than a year. A whole fleet, or a whole year: CCS's own 5 to 8 for one ship, which they do state is "during the whole year"; FORCE's 2 to 3 per cent, which they state as "in overall fleet operations"; and GreenVoyage2050's 0.5 to 2 for most of the world fleet, which IMO say is "derived from full-scale trials and detailed calculations carried out on a number of different ships in different routes". Only CCS use the word year. Reading the other two as year-weighted is our inference about what an average that broad must already contain, and it is worth saying whose inference it is. What none of the three does any longer is let one good condition carry the number on its own.

So carry this sentence and use it every time somebody quotes you a trim saving: before you ask how much, ask per cent of what. It is the single most useful question in the subject.

The two numbers need not even come from different studies. Take one ship, one trim table, one afternoon: the course ship, whose table the workbook carries. At her lightest regular condition, 8.0 metres and 14 knots, she needs 6,706 kW at the trim she habitually sails and 6,406 kW at the table's best trim. That is 300 kW, or 4.47 per cent: a true figure, and you may quote it.

Now weight it by the year she actually sails. Of 6,000 sea hours, 2,100 fall in that light band. Another 2,400 fall at 10.0 metres and 16 knots, where she sails half a metre by the stern at 12,016 kW and the table's best trim needs 11,536: 480 kW, or 3.99 per cent. The last 1,500 are deep, 12.0 metres at 18 knots, and there the table's best trim is the one she is already on: nothing saved, 0.00 per cent.

Weight by fuel, not by hours. An hour deep burns nearly three times what an hour light burns, and it is fuel that pays the bill. The energy saved is 300 × 2,100 plus 480 × 2,400 plus nothing, which is 1,782,000 kilowatt-hours a year. The energy burned is 6,706 × 2,100 plus 12,016 × 2,400 plus 18,685 × 1,500, which is 70,948,500. Divide one by the other: 2.51 per cent across the year.

Same hull, same table, same afternoon. The second is smaller than the first by a factor of 1.8, and it is the only one of the two that belongs in a budget. The course workbook does exactly this sum, in cells B87 to B91; doing it once by hand is how you learn what the sheet is doing. Average the three percentages by hours instead, 0.35 × 4.47 plus 0.40 × 3.99 plus 0.25 × nothing, and you get 3.16, a quarter too high, because that gives the deep, thirsty hours the same voice as the cheap light ones.

The spread is physics, not marketing

Our course Ship Resistance Fundamentals owns the calm-water decomposition, the bulbous bow tuned to one design speed and one design draught, and the form factor that operations selects rather than sets. This lesson takes that as given and says only this: a change of trim touches some of those terms and barely reaches others, which is lesson 2's entire subject.

That asymmetry is why the published range is so wide. GreenVoyage2050 states the consequence in one sentence, worth having verbatim: "full-body ships where resistance from viscous friction is higher than wave friction (e.g. tank and bulk) will generally have a less reduction by optimising the trim and draft". A fine container hull and a laden bulker are therefore not being told the same thing by the same percentage.

The screening sum, before you spend anything

The prize is not the saving at a condition. It is that saving multiplied by the fuel she burns near it, summed over the conditions she really visits: the sum worked above, and the only trim calculation available before any money moves.

Two features predict a large prize: a wide draught range actually used, because a table can only pay where the ship leaves the condition the hull was optimised for, and a lot of time at part load, which is the same statement in operational clothes and is why GreenVoyage2050's carve-out names container and ro-ro tonnage. Two predict none: a full hull form, for the reason quoted above, and a ship that trades at design draught and design speed most of the year, because a hull run at the condition it was optimised for has close to nothing left.

So do this first, out of the ship's own record. Take the last twelve months of logged draughts, the mean draught of each sea passage, and compute the share of annual sea hours spent below about 80 per cent of design draught. For this course's ship, whose design draught is 12.0 metres, that threshold is 12.0 × 0.8 = 9.6 metres. Her 2,100 sea hours at 8.0 metres fall below it and her other 3,900 do not, so that is 2,100 ÷ 6,000 = 35 per cent of her year in the band where a trim table has something to sell: she is a candidate.

The seven steps you would be buying

Seven steps stand between that fact and a policy anyone should sign. Which terms of the resistance a change of trim may touch, and why the mechanism everybody names is the wrong one (lesson 2). What trim she is actually sailing, as against the trim on the loading computer (lesson 3). Where her trim table came from: computed, towed, or regressed out of her own service data (lesson 4). Where that table stops being entitled to answer, and how to refuse (lesson 5). Whether its claim can be tested at all against the spread between prediction methods and the noise in-service measurement leaves behind (lesson 6). What the optimum costs in ballast and forgone cargo, because a trim reached with 1,800 tonnes of water is not free (lesson 7). And the larger lever underneath trim all along, draught and displacement (lesson 8).

The decision this lesson buys costs nothing, and you can make it this afternoon. Run the 80 per cent sum. If the answer is 400 hours out of 6,000, under seven per cent of her year, write that number down and stop: a trim programme on a ship that always sails deep produces a saving too small for anyone to demonstrate afterwards, and an unverifiable saving is a report rather than a result. If the answer is a third of her year, she is a candidate, and the rest of the course prices her.

Fix one habit alongside it. When a master, a charterer or the office asks what trim she should sail at, establish which of two questions is being asked: they never have the same answer. "What is best right now, at this draught and this speed" is a table look-up, and lessons 3 to 5 make that look-up safe. "What trim policy is worth money over a year" is the screening sum above, and it is the only one of the two with a business case attached.

Lesson 2 next, where the term everybody names first turns out not to be the one doing the work.