How to Conduct a Draft Survey: Step-by-Step Guide

A draft survey is the primary method for determining the weight of bulk cargo loaded onto or discharged from a vessel by measuring the ship's displacement before and after cargo operations. It is widely used in the international shipping industry for commodities such as coal, iron ore, grain, fertilizers, and aggregates. While the underlying principle — Archimedes' principle of buoyancy — is straightforward, executing an accurate draft survey requires meticulous attention to detail, calibrated instruments, and a thorough understanding of the corrections that must be applied to raw readings. This guide walks through every phase of the process, from the preparation stage through to the final survey report.

Table of Contents

1. Step 1: Preparation and Safety
2. Step 2: Taking Draft Readings
3. Step 3: Applying Corrections to Observed Drafts
4. Step 4: Determining Displacement from Hydrostatic Tables
5. Step 5: Calculating Cargo Weight
6. Step 6: Compiling the Survey Report
7. Common Mistakes to Avoid
8. Frequently Asked Questions

Step 1: Preparation and Safety

Before approaching the vessel, assemble and verify all equipment. Essential tools include a certified draft gauge or calibrated digital draft reader, a ballast water measuring tape with water-finding paste, a hydrometer or digital density meter for measuring harbor water density, a thermometer, a sounding tape for tank measurements, and the vessel's approved stability booklet containing the latest hydrostatic tables. Check that every instrument has a valid calibration certificate — using uncalibrated equipment is one of the most frequent causes of draft survey disputes. Review the vessel's general arrangement plan to identify the locations of all draft marks, ballast tanks, and sounding pipes. Conduct a safety briefing with the survey team covering personal protective equipment, communication protocols with the vessel crew, and emergency procedures. Verify that sea conditions are within acceptable limits: swell height below 0.5 meters and wind speed under 15 knots are generally considered acceptable thresholds for accurate readings. If the harbor water density is expected to vary significantly with depth, plan to take density samples at multiple levels using a water sampler.

Step 2: Taking Draft Readings

Draft readings must be taken at six standard positions: forward port, forward starboard, midship port, midship starboard, aft port, and aft starboard. At each position, read the draft mark where the water surface intersects the scale. In wavy conditions, take at least three readings — one at a crest, one at a trough, and one at the mean level — and average them. Record readings to the nearest centimeter on metric marks or half-inch on imperial marks. Never estimate beyond the precision of the marks themselves. Simultaneously, measure the harbor water density by lowering a bucket or sampling device to approximately half the vessel's draft and testing the sample immediately with a calibrated hydrometer. The density of seawater typically ranges from 1.020 to 1.030 g/cm³ depending on salinity and temperature; harbor water can be significantly lower, especially near river mouths. Record the density at each survey — initial and final — because changes in water density between surveys directly affect displacement calculations. GOTEC's AI draft reading system can capture simultaneous port and starboard readings using stabilized cameras and computer vision, eliminating the need for manual averaging and dramatically reducing the time spent at each mark position.

Step 3: Applying Corrections to Observed Drafts

Raw draft readings are almost never directly usable. A series of corrections must be applied to obtain the true mean draft. First, correct for trim — the difference between forward and aft drafts. When a vessel is trimmed, the forward and aft draft marks are not located at the perpendiculars; the observed values must be corrected to the forward perpendicular (FP) and aft perpendicular (AP) using the vessel's trim correction table. Second, apply a heel or list correction: if the vessel is not upright, the mean of port and starboard readings does not equal the centerline draft. For small angles of heel (under 2 degrees), a simple average is acceptable; for larger angles, a trigonometric correction using the vessel's beam is required. Third, apply a deflection correction if the vessel is hogging (ends deeper than midship) or sagging (midship deeper than ends). The mean of means draft is calculated as: (Forward Mean + 6 x Midship Mean + Aft Mean) / 8. This weighted average accounts for the parabolic deflection curve of most loaded vessels. Finally, correct for water density: the displacement obtained from hydrostatic tables assumes a specific density, typically 1.025 g/cm³ for seawater. For each 0.001 deviation from the table density, displacement changes by approximately 0.1%.

Step 4: Determining Displacement from Hydrostatic Tables

With the corrected mean draft, enter the vessel's hydrostatic tables to find the corresponding displacement. Hydrostatic tables are specific to each vessel and are contained within the stability booklet approved by the classification society. For drafts falling between tabulated values, linear interpolation is standard practice: displacement = D1 + [(D2 - D1) x (draft - T1) / (T2 - T1)], where D1 and D2 are the tabulated displacements at drafts T1 and T2. Always double-check trim corrections to displacement — many hydrostatic tables include separate trim correction values, termed the "first trim correction" and "second trim correction." The first trim correction adjusts the longitudinal center of buoyancy (LCB) position; the second adjusts for changes in the moment to change trim. For vessels with significant trim (exceeding 1% of length between perpendiculars), neglecting trim corrections can introduce errors of 50 to 200 tonnes even on medium-sized bulk carriers. If the vessel is listed, apply a heel correction to displacement as well. Modern digital tools have streamlined this step considerably: integrated draft survey software can accept raw readings as input and automatically interpolate hydrostatic values, applying all corrections in sequence. This is consistent with the broader trend toward digitization across maritime operations.

Step 5: Calculating Cargo Weight

The fundamental draft survey equation is: Cargo Weight = (Loaded Displacement – Light Displacement) – (Total Consumable Changes). Loaded displacement is the displacement after cargo operations are complete (final survey); light displacement is the displacement before cargo operations begin (initial survey). The difference is the total weight change of the vessel — but not all of this change is attributable to cargo. Consumables — fuel oil, diesel oil, lubricating oil, fresh water, ballast water, stores, and crew effects — all change between surveys. Each consumable quantity must be measured at both the initial and final surveys via tank soundings, and the net change in weight must be deducted from the displacement difference. Ballast water is frequently the largest variable; a thorough ballast measurement protocol (see our guide to measuring ballast water accurately) is essential. The final cargo weight should be expressed in metric tonnes to the nearest 0.1 tonne. It is good practice to calculate the draft survey accuracy by comparing the result against shore scale figures when available and investigating any discrepancies exceeding 1.0%.

Step 6: Compiling the Survey Report

A complete draft survey report serves as a legal document that may be scrutinized in cargo quantity disputes, insurance claims, and charter party settlements. The report must include: vessel name, IMO number, port, berth, date and time of each survey, surveyor name and credentials, environmental conditions (sea state, wind, visibility), all raw draft readings at six positions with timestamp, the corrections applied (trim, heel, deflection, density) with intermediate values, the corrected mean draft, the displacement from hydrostatic tables with interpolation notes, a complete inventory of all consumables at initial and final surveys with sounding records, the density measurements at each survey, and the final calculated cargo weight. Attach photographs of all draft marks — timestamped and labeled — as supporting evidence. An increasing number of survey companies are adopting digital reporting platforms that embed AI-validated images and automated calculations, creating an auditable chain of custody for every data point. GOTEC's integrated reporting module generates these reports automatically from field data, reducing reporting time from approximately 45 minutes to under 10 minutes per survey.

Common Mistakes to Avoid

Several recurring errors compromise draft survey accuracy. Reading draft marks from the wrong angle — parallax error can introduce 2 to 5 cm of error per reading if the surveyor's eye is not level with the water surface. Always read marks from as close to water level as possible. Neglecting ballast water measurement — incomplete ballast tank soundings are the single largest source of draft survey error, potentially misrepresenting cargo weight by hundreds of tonnes. Sound every tank, not just those declared as full or empty. Using the wrong hydrostatic data — always confirm that the stability booklet onboard is the latest edition approved for the vessel after any structural modifications. Failing to account for consumable changes — even minor fuel oil consumption during a 24-hour loading operation can affect results by 5 to 15 tonnes. Density sampling at the surface only — harbor water density often stratifies, with fresh river water sitting atop denser seawater. A sample taken at half the vessel's draft is the minimum standard. Incomplete documentation — without timestamped photographs and raw data logs, a survey report may not withstand scrutiny in a formal dispute.

Frequently Asked Questions

What is the minimum acceptable accuracy for a draft survey?

Industry standards generally accept a draft survey accuracy of 0.5% to 1.0% of the total cargo weight when conducted by a qualified surveyor under favorable conditions. For a Panamax bulker carrying 75,000 tonnes of iron ore, this translates to an acceptable tolerance range of approximately 375 to 750 tonnes. In practice, experienced surveyors using calibrated equipment and digital tools routinely achieve results within 0.3% to 0.5%. Discrepancies between draft survey and shore scale figures exceeding 1.0% should trigger a full review of methodology, equipment calibration, and consumable measurements. For more on the underlying technology, see our overview of GOTEC draft survey equipment.

How has AI changed the draft survey process?

AI has transformed draft surveying in three key areas. First, automated draft reading uses stabilized camera systems with computer vision to read draft marks at all six positions simultaneously, eliminating parallax error, reducing reading time from approximately 20 minutes to under 2 minutes, and producing a digital record that can be independently verified. Second, automated correction calculations integrate the vessel's hydrostatic tables, trim correction formulas, and density adjustments into a single software pipeline, eliminating manual calculation errors. Third, predictive analytics can flag potential discrepancies — such as an unexpected draft change inconsistent with reported cargo quantity — before the survey is finalized. For ports processing multiple vessels per day, the cumulative time savings are measured in hours per vessel per call. Learn more about how AI extends into customs inspection workflows.