How to do a draft survey calculation, step by step
A draft survey weighs a ship by Archimedes' principle. You read the drafts, find the displacement in the ship's hydrostatic tables, correct it for trim and for the density of the water she is floating in, and subtract everything on board that is not cargo. Done once before loading and once after (or before and after discharge), the difference is the cargo.
It is the standard commercial method for bulk cargoes such as coal, ore and grain, and its figure often goes on the bill of lading. Chief officers do it alongside the attending surveyor so that the ship has her own figure. Because money rides on the result, the arithmetic is done in a fixed sequence, the one set out in the UN ECE Code of Uniform Standards and Procedures for the Performance of Draught Surveys of Coal Cargoes (1992) and in the standard texts on bulk carrier practice.
This guide follows that sequence with the formulas, then works a complete loading survey with every correction.
What you need before you start
- The hydrostatic tables: displacement, TPC (tonnes per centimetre immersion), LCF (longitudinal centre of flotation) and MCTC (moment to change trim one centimetre) against draft, and the water density the tables are drawn up for, usually 1.025 t/m³.
- The length between perpendiculars (LBP) and the distance of each set of draft marks from its perpendicular or from midships. These are in the stability booklet or on the capacity plan.
- The lightship weight.
- A hydrometer and a sampling can for the dock water.
- Soundings of every ballast, fresh water and fuel tank, with the tank tables.
Sign conventions matter more here than anywhere else in the calculation. This guide uses the same ones as the draft survey calculator: trim is positive by the stern, the LCF is measured from midships and is positive forward, and mark distances are positive when the marks lie inboard of the perpendiculars.
Step 1: read the six drafts
Read the marks forward, midships and aft, on both sides: six readings. Take the offshore side from a boat if you can, as close to the waterline as possible. In any ripple or swell, watch several wave cycles and take the mean of the crests and troughs rather than a single glance. Read port and starboard at the same station close together in time, because the ship moves as cranes swing and ballast runs.
Average each pair:
F = (F_port + F_stbd) ÷ 2 M = (M_port + M_stbd) ÷ 2 A = (A_port + A_stbd) ÷ 2
The port and starboard mean removes a small list from the figure, but only approximately. If the two sides differ by more than about 10 cm, bring the ship upright before reading if at all possible.
Step 2: correct the drafts to the perpendiculars
The hydrostatic tables assume drafts at the forward perpendicular (FP), the aft perpendicular (AP) and midships. The marks are painted where the hull shape allows, usually some metres inside the perpendiculars. When the ship is trimmed, the draft at a mark is not the draft at its perpendicular, so each reading is slid along the slope of the waterline:
Apparent trim t_a = A − F
LBM = LBP − d_F − d_A length between the marks
F_FP = F − d_F × t_a ÷ LBM d_F: forward marks aft of the FP
A_AP = A + d_A × t_a ÷ LBM d_A: aft marks forward of the AP
M_mid = M − d_M × t_a ÷ LBM d_M: midship marks aft of midships
True trim t = A_AP − F_FP
With a stern trim and marks inboard of the perpendiculars, the forward draft gets smaller and the aft draft larger, so the true trim is always a little more than the apparent trim. If your corrections make it smaller, a sign is wrong.
Step 3: the mean of means
A ship is not rigid. Loaded heavily amidships she sags, loaded at the ends she hogs, and the mean of the forward and aft drafts then misrepresents the average immersion. The survey uses a weighted mean that gives the midship draft six parts in eight:
Mean of ends = (F_FP + A_AP) ÷ 2
Mean of means = (mean of ends + M_mid) ÷ 2
Quarter mean D_qm = (mean of means + M_mid) ÷ 2 = (F_FP + A_AP + 6 × M_mid) ÷ 8
Hog or sag = M_mid − (F_FP + A_AP) ÷ 2 (+ sag, − hog)
Enter the hydrostatic tables with D_qm and take out the displacement, the TPC and the LCF, interpolating between the tabulated drafts. Also take the MCTC at D_qm + 50 cm and at D_qm − 50 cm for the second trim correction. If you want to see where TPC and MCTC come from, the hydrostatics calculator builds them from the hull dimensions.
Step 4: first and second trim corrections
The tables are for a ship on even keel. A trimmed ship pivots about her centre of flotation, not about midships, so the draft that truly corresponds to the displacement is the draft at the LCF. The first trim correction converts the difference between the two into tonnes:
FTC = trim (cm) × LCF distance from midships (m) × TPC ÷ LBP
The sign rule is that the correction is positive when the LCF lies towards the deeper end. Trimmed by the stern with the LCF aft of midships: add. Trimmed by the stern with the LCF forward of midships: subtract. With the conventions above, that is FTC = −t × LCF × TPC ÷ LBP with t in centimetres. The LCF moves with draft, so the sign can change between the initial and final surveys, as it does in the example below.
The second trim correction, often called the Nemoto correction, allows for the LCF itself shifting as the trimmed waterplane changes shape:
STC = 50 × t² × (MCTC at D_qm + 50 cm − MCTC at D_qm − 50 cm) ÷ LBP
with t in metres. It is always added, whichever way the ship is trimmed. It grows with the square of the trim: negligible near even keel, a few tonnes at a metre, and tens of tonnes at the large trims of a ballast condition.
Δ_trim = Δ_table + FTC + STC
Step 5: density correction
The tables give the mass displaced in water of the table density. In lighter dock water the same underwater volume weighs less:
Δ_corrected = Δ_trim × ρ_dock ÷ ρ_table
Draw the sample at about mid-draft, on both sides of the ship and clear of overboard discharges, and read the hydrometer straight away. Use a hydrometer that reads on the same basis as the tables. The density can change with the tide and with depth, so take it at the time of the draft readings, not an hour before.
Step 6: deductibles, lightship and the constant
From the corrected displacement, take away everything measurable that is not cargo:
Deductibles = ballast + fresh water + fuel oil + diesel oil + lubricating oil + others
Net = Δ_corrected − deductibles − lightship
Cargo = Net − constant
Each tank is sounded, the sounding corrected for trim and list, converted to volume with the tank tables and then to mass with the density of what is in it. A sounding pipe in the aft corner of a tank reads deep when the ship is trimmed by the stern; the tank sounding correction calculator shows the size of that effect, and the bunker quantity calculator converts fuel volume and temperature to tonnes. Measure the ballast density rather than assuming it, particularly if the ballast was taken in a river or changed at sea by ballast water exchange.
The constant is whatever remains: stores, spares, crew and effects, paint, mud in the ballast tanks, and every small error in the lightship and the tables. It is found at the survey where no cargo is on board, by setting the constant to zero; the net figure is then the constant. For a bulk carrier it is typically 200 to 500 t and should be much the same from one port to the next.
Worked example: a loading survey
A bulk carrier has LBP 185 m. The forward marks are 2.4 m aft of the FP, the aft marks 7.5 m forward of the AP and the midship marks 0.8 m aft of midships. Lightship is 10 150 t and the tables are for 1.025 t/m³.
Initial survey, in ballast
Readings: forward 4.80 m port and 4.84 m starboard, midships 5.52 and 5.60 m, aft 6.20 and 6.24 m. Dock water 1.018 t/m³.
- Means: F = 4.82 m, M = 5.56 m, A = 6.22 m. Apparent trim 6.22 − 4.82 = 1.40 m by the stern.
- LBM = 185 − 2.4 − 7.5 = 175.1 m, so the slope is 1.40 ÷ 175.1 = 0.007995 m per metre.
- F_FP = 4.82 − 2.4 × 0.007995 = 4.801 m. A_AP = 6.22 + 7.5 × 0.007995 = 6.280 m. M_mid = 5.56 − 0.8 × 0.007995 = 5.554 m.
- True trim: 6.280 − 4.801 = 1.479 m by the stern. Sag: 5.554 − 5.540 = 1.3 cm.
- Quarter mean: (4.801 + 6.280 + 6 × 5.554) ÷ 8 = 5.550 m.
- Tables at 5.550 m: displacement 28 420 t, TPC 46.8 t/cm, LCF 4.2 m forward of midships, MCTC 412 t·m/cm at +50 cm and 398 at −50 cm.
- First trim correction: 147.9 × 4.2 × 46.8 ÷ 185 = 157.2 t. The LCF is forward and the ship is deeper aft, so it is subtracted: −157.2 t.
- Second trim correction: 50 × 1.479² × (412 − 398) ÷ 185 = +8.3 t.
- Displacement corrected for trim: 28 420 − 157.2 + 8.3 = 28 271.1 t.
- Density: 28 271.1 × 1.018 ÷ 1.025 = 28 078.0 t.
- Deductibles: ballast 16 640, fresh water 210, fuel oil 640, diesel oil 85, lubricating oil 40, others 15, total 17 630 t.
- Net: 28 078.0 − 17 630 − 10 150 = 298.0 t. No cargo is on board, so this is the constant.
Final survey, loaded
Readings: forward 10.86 and 10.90 m, midships 11.02 and 11.10 m, aft 11.20 and 11.26 m. Dock water 1.021 t/m³. The same steps give:
| Item | Initial | Final |
|---|---|---|
| Mean drafts at the marks F / M / A (m) | 4.82 / 5.56 / 6.22 | 10.88 / 11.06 / 11.23 |
| Drafts at FP / midships / AP (m) | 4.801 / 5.554 / 6.280 | 10.875 / 11.058 / 11.245 |
| True trim by the stern (m) | 1.479 | 0.370 |
| Quarter-mean draft (m) | 5.550 | 11.059 |
| Table displacement (t) | 28 420 | 55 960 |
| TPC (t/cm); LCF from midships (m) | 46.8; 4.2 forward | 53.4; 2.1 aft |
| MCTC at +50 cm / −50 cm (t·m/cm) | 412 / 398 | 520 / 498 |
| First trim correction (t) | −157.2 | +22.4 |
| Second trim correction (t) | +8.3 | +0.8 |
| Displacement corrected for trim (t) | 28 271.1 | 55 983.2 |
| Dock-water density (t/m³) | 1.018 | 1.021 |
| Displacement corrected for density (t) | 28 078.0 | 55 764.8 |
| Deductibles (t) | 17 630 | 1 084 |
| Lightship (t) | 10 150 | 10 150 |
| Net (t) | 298.0 | 44 530.8 |
The final deductibles are ballast 180 t (unpumpable remains), fresh water 160, fuel oil 610, diesel oil 80, lubricating oil 39 and others 15. In the final survey the LCF is aft and the ship is still by the stern, so the first trim correction is added: 37.0 × 2.1 × 53.4 ÷ 185 = 22.4 t.
Cargo loaded = final net − constant = 44 530.8 − 298.0 = 44 232.8 t. The same figure comes from the difference of the two displacements less their deductibles: (55 764.8 − 1 084) − (28 078.0 − 17 630) = 54 680.8 − 10 448.0 = 44 232.8 t. Lightship and constant cancel, which is why a survey by difference does not depend on either being exactly right.
The draft survey calculator reproduces both columns: enter the initial survey with the constant at 0, then the final survey with the constant at 298.0 t.
What accuracy to expect
Under good conditions a draft survey is typically within about ±0.5 % of the cargo. On the example that is roughly 220 t, the equivalent of about 4 cm on the mean draft, since one centimetre of quarter-mean draft is one TPC: 53.4 t at the loaded draft. The reading of the marks dominates the error, and these are the things that make it worse:
- Swell and chop. The waterline moves up and down the marks and the eye favours the crests. Average over several cycles, and read in the lee of the hull or a boat where you can.
- Hog and sag. The quarter mean assumes the hull bends in a smooth curve. A deflection beyond about 10 cm stretches that assumption and should be reported with the result.
- List. The mean of the two sides is only an approximation for a listed ship, and the list also upsets every tank sounding.
- Large trim. Above about 1 % of LBP both trim corrections become large and the even-keel tables lose accuracy; use trimmed hydrostatics if the ship has them. Shifting ballast to reduce the trim before the survey helps, and the trim change calculator tells you how much to move.
- Silt and mud in ballast tanks. Sediment sits below the sounding and weighs more than water. It ends up in the constant at one survey and may be pumped out, or not, before the next.
- Wrong density sampling. A surface sample near an outfall or after rain, a sample taken long before the readings, or a hydrometer on the wrong basis each scale the whole displacement.
- Ballast quantities. In a loading survey the initial ballast is the largest deductible by far (16 640 t above), so a small percentage error in it outweighs everything else. Sound every tank and check that "full" tanks are overflowing from all their air pipes.
- The ship touching bottom, or a tight mooring or strong current holding her down. The displacement read is then not her weight.
Common mistakes
- Wrong sign on the first trim correction. Check where the LCF is at each survey; do not carry the sign over from the previous one.
- Tables that give the LCF from the aft perpendicular. Convert it to a distance from midships before using the formula: subtract half the LBP.
- Mixing centimetres and metres. The first correction takes trim in centimetres, the second in metres.
- Using the apparent trim, from the marks, in the trim corrections instead of the true trim between perpendiculars.
- Entering the tables with the mean of the ends or the midship draft instead of the quarter mean.
- Taking MCTC at the quarter-mean draft only. The second correction needs the difference between the values 50 cm above and 50 cm below.
- Applying the density correction before the trim corrections, or to the cargo instead of to the displacement.
- Assuming 1.025 for ballast. Ballast taken in a river port is lighter.
- Using the constant from the ship's papers without question. If the measured constant comes out negative, or far from its usual value, look for an error in the drafts or the tanks before accepting either figure.
- Letting operations run on during the survey. Ballast, bunkering, hatch covers and cranes must be stopped while drafts and soundings are taken.
Limits of the method
The calculation is only as good as the hydrostatic tables, which describe the ship as built and on even keel. It corrects for trim and, through the quarter mean, for moderate hull deflection; it does not correct for list, for a hull deflected in an irregular shape, or for errors in the tank tables. The surveyor's judgement of sea state and reading conditions is part of the result and should be written on the report.
Draft surveys suit bulk cargoes, where the cargo is a large share of the displacement. Tankers measure cargo by ullage and container ships work from declared weights, although the same arithmetic gives a useful cross-check of the deadweight on any ship. For the terms used here, see the maritime glossary.
Calculators used in this guide
- MaritimeDraft surveySix observed drafts to corrected drafts, quarter-mean draft, trim and density corrections, net displacement and cargo or constant from the hydrostatic tables.
- MaritimeHydrostatics basicsDisplacement from draft and block coefficient, TPC, fresh and dock water allowance, trim and end drafts from LCG, LCB and MCTC.
- MaritimeTank sounding trim and heel correctionCorrected sounding or ullage of a rectangular tank for trim and heel at an off-centre sounding pipe, with volume, mass and the wedge check.
- MaritimeBunker quantity and density correctionBunker mass in vacuo and in air from sounded volume, temperature and density at 15 °C (Table 54B VCF), with energy content, cost and BDN comparison.
- MaritimeChange of trim and new drafts calculatorNew forward and aft drafts after shifting, loading or discharging a weight, from MCTC, TPC and LCF: bodily sinkage, change of trim and draft changes.
- MaritimeBallast water exchange timeD-1 ballast exchange: sequential, flow-through and dilution times and distance, the 3-volume / 95 % rule from perfect mixing, and the location criteria.
Results are estimates. They do not replace the governing code, class or flag requirements, or a check by a competent person. Report an error in this guide
More guides
- Cargo securing calculation to CSS Code Annex 13
- How EU ETS and FuelEU Maritime costs are calculated
- How to calculate a ship's CII rating, step by step
- Hull and propeller performance in service with ISO 19030
- Ship resistance and powering: from hull to installed power
- How to calculate ship squat and under-keel clearance