What is a draft survey? Method, formulas and worked example

A draft survey determines the weight of a bulk cargo from the ship's draft. The surveyor reads the draft marks, measures the dock water density and sounds the tanks, once before cargo work starts and once after it ends. The ship's hydrostatic tables turn each set of readings into a displacement; the difference between the two, net of ballast, fuel and water, is the cargo.

When a draft survey is carried out

The initial survey is taken immediately before loading or discharge begins, the final survey immediately after it ends, in loading and discharge ports alike. Under the UNECE code for draught surveys of coal cargoes (1992), no cargo should be worked until the surveyor has finished the initial readings and given consent, and a licensed deck officer witnesses the drafts. The code treats the resulting report as equivalent to a weight certificate: see the draft survey report and weight certificates.

How a draft survey is calculated, step by step

The sequence follows that code (ECE/ENERGY/19) and our own draft survey form. Trim is positive by the stern; the LCF (centre of flotation) is positive aft of midships.

1. Six drafts, forward, midships and aft on both sides, averaged per pair: Tf, Tm, Ta. Apparent trim = Ta − Tf.
2. Perpendiculars. LBM = LBP − Xf − Xa, with Xf the fore marks aft of the forward perpendicular and Xa the aft marks forward of the aft perpendicular. Tfp = Tf − (apparent trim × Xf) / LBM; Tap = Ta + (apparent trim × Xa) / LBM; true trim = Tap − Tfp.
3. Quarter mean = (mean of means + Tm) / 2, where mean of means = (Tfp + Tap + 2 × Tm) / 4. It is the code's approximate allowance for hog and sag.
4. Hydrostatic tables at the quarter mean: displacement D(T), TPC (tonnes per cm immersion), LCF, and MTC (moment to change trim 1 cm) 50 cm above and below.
5. First trim correction = 100 × true trim × LCF × TPC / LBP, added with its sign.
6. Second trim correction (Nemoto) = 50 × true trim² × (MTC₊₅₀ − MTC₋₅₀) / LBP, always positive. Approved trim correction tables replace both formulas.
7. Density correction = displacement corrected for trim × (ρ − 1.025) / 1.025, with ρ the dock water density from a hydrometer calibrated in air. For load line purposes, see fresh water and dock water allowance.
8. Deductibles: ballast, fresh water, bunkers, lubricating oil, bilges and slops, all sounded at the initial survey; at the final survey the code re-sounds the ballast and lets fuel and fresh water follow from port consumption. What remains is the net displacement.
9. Cargo = net displacement final − net displacement initial.

Our form assumes fore and aft marks inside the perpendiculars (enter Xf or Xa as negative otherwise), midship marks halfway between them and tables for 1.025 t/m³. The code corrects for midship marks elsewhere and uses the ship's own reference density.

Worked example: soyabean meal loaded in Rotterdam

A fictitious vessel of 180.00 m LBP, with Xf 3.00 m and Xa 5.00 m, so LBM is 172.00 m. Dock water 1.0210 t/m³ at both surveys. The hydrostatic figures are illustrative.

	Initial survey	Final survey
Drafts fwd, mid, aft, port/stbd (m)	4.18/4.22, 5.58/5.62, 6.98/7.02	10.98/11.02, 11.28/11.32, 11.58/11.62
Tfp / Tap (m)	4.1512 / 7.0814	10.9895 / 11.6174
True trim (m)	2.9302	0.6279
Quarter mean draft (m)	5.604	11.301
D(T) / TPC / LCF	21,450 MT / 32.5 MT/cm / +2.1 m	46,800 MT / 36.8 MT/cm / +1.4 m
MTC at +50 / −50 cm (MT·m/cm)	285 / 272	512 / 498
First / second trim correction	+111.105 / +31.006 MT	+17.972 / +1.533 MT
Displacement corrected for trim	21,592.111 MT	46,819.505 MT
Density correction	−84.262 MT	−182.710 MT
Displacement corrected for density	21,507.849 MT	46,636.795 MT
Total deductibles	9,305.000 MT	1,104.000 MT
Net displacement	12,202.849 MT	45,532.795 MT

Cargo = 45,532.795 − 12,202.849 = 33,329.946 MT. Against a bill of lading weight of 33,400.000 MT, the difference is −70.054 MT, or −0.210%.

Steps are computed unrounded and shown rounded, so a line recomputed from the table can differ in the last decimal (initial first trim correction: 100 × 2.930233 × 2.1 × 32.5 / 180 = 111.105 MT).

A weighing by difference

The cargo is the difference between two net displacements, so an error in either survey lands on it in full. In the example, by our calculation:

• One centimetre on the final quarter mean draft is roughly the TPC, 36.8 MT: 0.110% of this cargo, but 0.736% of a 5,000 MT parcel on the same ship.
• An error of 0.001 t/m³ in the final density moves the result by 45.678 MT.
• A 1% error in the 8,450 MT of initial ballast, 84.500 MT, exceeds the whole difference with the bill of lading.

Why ballast tanks are sounded, not taken from the log

The code wants every ballast tank sounded with a calibrated steel tape and water-finding paste, and does not accept "full" or "empty" as an entry. A pressed-up tank can hold air pockets when the ship is trimmed; a tank that reads dry at an aft sounding pipe can hold water forward when she trims by the head. Ballast density varies between fresh and sea water, so it is measured too.

How accurate is a draft survey?

The UNECE code states that following it, with the ship's class-approved draught survey manual and qualified, experienced surveyors, gives cargo weights consistently within 0.5%; checks on the same ship have reached 0.1% in favourable conditions. The West of England P&I Club puts the typical variation at 0.5 to 1.0%.

On this cargo 0.5% is 166.650 MT, so a difference of −70.054 MT neither proves nor rules out a shortage. Where the bill of lading figure comes from a shore scale, Britannia advises against reconciling the two: they are different methods, and the charter party should say which figure binds. A shortage that holds up is covered in cargo claims in bulk shipping.

Where published methods disagree

Published sources differ on two terms. Our form makes a choice on both; agree yours with the other side's surveyor beforehand.

The sign of the first trim correction. The UNECE code uses a sign table: trim by the stern with the LCF aft adds, with the LCF forward subtracts, and the reverse for trim by the head. Britannia notes that tables give the LCF from midships or from the aft perpendicular, and that signs differ between ships. Our form enters the LCF with a sign, positive aft, and adds the correction, which matches the UNECE table. With the sign wrong at both surveys, the example cargo becomes 33,515.485 MT, some 185.5 MT more.

The form of the second trim correction. The UNECE code uses the difference between the MTC 50 cm above and 50 cm below the mean draft; at least one published guide uses a single MTC read at the quarter mean. In the example the MTC values are 272 to 512 MT·m/cm and their differences 13 and 14, so the results differ widely. Our form uses the code's difference, read around the quarter mean.

When a draft survey is the right method

A draft survey is done only on ships, by the draught, and only when there is no other way to determine the weight. Containers and road trucks always go over a weighbridge, and for liquids the shore tank measurement is the determination of weight, so a draft survey belongs to dry bulk. Surveyors who work mainly on vegetable oil rarely do one; the calculation needs the ship's dimensions and differs from ship to ship.

The UNECE code was drawn up for coal cargoes but recommends its procedure for dry cargoes in general. Like any quick method for weighing large masses, it has limits. It serves less well for a parcel small relative to the ship, at a berth exposed to swell or a current strong enough for squat (the code sees no effect below 2.5 knots), with trim beyond the range of the tank tables' trim corrections (the surveyor then stops until the ship is re-trimmed), and without the approved hydrostatic data on which the 0.5% depends.

FOSS

FOSS is software for inspection companies and the port, with a sector specialist at the table. Readings are entered once, offline on board if need be, and nine report types come from that input. An outturn more than 0.5% short of the bill of lading weight flags a protest automatically.