BRIDGE → Stability & Cargo
Position on the Bridge
System Group: Stability Management / Cargo Planning
Primary Role: Real-time calculation of hull stress and stability parameters based on declared loading conditions
Interfaces: Cargo plans, draught surveys, tank sounding systems, port state control documentation, chief officer’s stability workbook
Operational Criticality: High — errors in loadicator output underpin every structural and stability decision made during loading and passage
Failure Consequence: Miscalculated GM or bending moment accepted as valid; ship departs in a condition outside permissible limits; structural damage or capsize without prior warning from any monitoring system
The instrument shows what was entered into it.
It knows nothing of what the sea is preparing to do.
Introduction
A loadicator is an approved computing tool. It is not a stability system in any comprehensive sense. The distinction matters every time a ship leaves port.
The instrument performs a specific, bounded calculation: given a declared distribution of weights, it returns still-water bending moments, shear forces, torsional loads where applicable, a metacentric height, and a righting lever curve. Those outputs are valuable. They are also a snapshot of a theoretical condition that existed at the moment the data was entered and has not existed since.
The problem is not the loadicator. The problem is the habit, common across the industry, of treating a green readout as clearance. An acceptable output confirms that the entered loading condition, if accurate, falls within the vessel’s approved limits. It confirms nothing else. The sea does not consult the loadicator before applying a sagging moment to a loaded bulk carrier in a head swell, and the loadicator has no sensor, no feed, and no awareness that the sea exists at all.
This article addresses what the loadicator calculates, the approval framework that gives those calculations legal standing, the ways in which input errors corrupt the output silently, and the failure modes that no stability computer on any ship in service can detect.
Contents
- 1. What the loadicator actually calculates
- 2. The approval framework and why it has operational consequences
- 3. Input integrity: the silent corruption problem
- 4. What the loadicator cannot calculate
- 5. The GM figure and its real-world limits
- 6. Righting arm curves and the conditions they assume
- 7. Cross-checking: draught, trim, and the physical ship
- 8. Closing Reality
1. What the loadicator actually calculates
The loadicator works from the hydrostatic data of the ship’s approved stability booklet, encoded at build or refit. Against that fixed hull geometry, it applies the weight distribution entered by the officer — cargo holds, tanks, stores, constants — and performs a numerical integration to produce the still-water bending moment (SWBM) and still-water shear force (SWSF) envelope along the ship’s length.
On vessels where torsional loading is a design concern — container ships and some RoRo types in particular — an approved loadicator will also calculate the still-water torsional moment. On bulk carriers and tankers, that calculation is rarely present or required.
From the same weight distribution, the system calculates displacement, LCG, TCG, KG, and from KG and the hydrostatics, the metacentric height GM. The full GZ curve is generated by applying the KG to the ship’s cross-curves of stability, producing righting levers at successive angles of heel.
These are not approximations. The calculations are precise given their inputs. The precision is real and the precision is entirely conditional on input accuracy. Both of those facts must be held simultaneously.
2. The approval framework and why it has operational consequences
A loadicator must be approved by the flag state administration or a recognised organisation acting on its behalf. That approval is not a rubber stamp. The approving body verifies that the hydrostatic data encoded in the system matches the stability booklet, that the permissible limit curves for SWBM and SWSF are correctly implemented, and that the output format allows the user to identify exceedances clearly.
Approval is vessel-specific. A system approved for one ship cannot be transferred to another, even of the same class, without separate approval. After structural alterations, after significant changes to the lightship condition, or after any modification that affects the hydrostatic tables, the approval lapses and must be reissued.
In port, PSC surveyors and terminal inspectors are entitled to ask for the loadicator’s approval certificate. They are also entitled to ask whether it has been updated following the ship’s last drydock. A loadicator running on lightship data that no longer reflects the actual ship is not an approved instrument in any meaningful sense, regardless of what the certificate says.
The approval certificate does not validate the current loading condition. It validates the instrument. The instrument is only as good as the data inside it and the data entered into it.
3. Input integrity: the silent corruption problem
Every error introduced at the input stage is amplified through the output. The loadicator has no mechanism to detect that the density entered for a fuel oil tank is wrong, that a hold has been declared empty when it contains residual cargo, or that the constant used for deck equipment and crew effects has not been updated since the ship’s last special survey added two tonnes to the lightship. It receives numbers and calculates. It does not audit.
Density errors are endemic. A forward ballast tank entered at 1.025 when the port water is 1.010 will produce a calculated KG that is fractionally but consistently wrong. Across multiple tanks, these errors accumulate in a direction that is not always predictable. On a vessel loading in a river estuary with a salinity gradient, the difference between declared and actual water density can shift the GM by several centimetres. That is not irrelevant on a ship already operating near its minimum GM requirement.
Lightship constants deserve particular attention. The constant is the catch-all for items not individually accounted for: crew and effects, provisions, spare gear, water in pipelines, residual liquids. It is estimated at the last inclining experiment and in practice drifts upward over the ship’s life as equipment is added and residuals accumulate. An understated constant produces an overstated GM. The loadicator will display a number that looks healthy. The actual metacentric height is lower.
Free surface correction is another persistent source of error. The loadicator applies a free surface moment based on the declared filling level of each liquid tank. If the officer enters a tank as pressed full when it is in fact 97% full, the free surface correction is zero. In reality there is a sloshing surface. The effect is small on a large tank pressed nearly full, but the habit of entering tanks as full or empty rather than their actual sounding — because it simplifies the entry — introduces a systematic bias toward optimism.
A loadicator output is only as honest as the officer who populated it.
4. What the loadicator cannot calculate
The loadicator calculates still-water conditions. The sea is not still.
Wave-induced bending moments — the hogging and sagging loads imposed by the wave profile passing along the hull — are entirely outside the system’s scope. Classification societies publish permissible total bending moment limits that are the sum of the still-water and wave-induced components. The loadicator only shows the still-water portion. The officer must understand that running close to the permissible SWBM limit in port leaves little margin for the wave-induced component once the ship enters a seaway. The loadicator will not say so.
Slamming loads, generated when the bow re-enters after emerging from a wave trough, produce enormous local structural stresses in the forward sections. These are transient, impulsive, and entirely dynamic. No shore-based stability calculation reaches them.
Green water loading — the weight of solid water shipped over the bow or deck in heavy weather — acts as a distributed added mass that changes trim, immerses the freeboard deck, and degrades intact stability. The loadicator knows nothing of it. The ship will behave differently than the GZ curve predicts.
Ice accretion is a static load, but one applied to exposed topsides, rigging, and deck equipment in a distribution that cannot be entered into a standard loadicator without deliberate manual input. That manual input is rarely made in real time, because the ice is forming in conditions where the officer has other priorities. The standard allowance figures in the stability booklet are a regulatory convenience, not a guarantee.
Parametric rolling is a resonance phenomenon that can drive roll amplitudes well beyond those suggested by the GZ curve, particularly on container vessels in following or quartering seas. The loadicator calculates the GZ curve from static hydrostatics. Parametric excitation is a dynamic interaction between the ship’s natural roll period, wave encounter frequency, and the variation in waterplane area as the vessel pitches. The loadicator is constitutionally blind to it.
The gap between what the loadicator shows and what the sea imposes is not a flaw in the instrument. It is the instrument’s fundamental nature. The flaw is in forgetting that nature exists.
5. The GM figure and its real-world limits
The GM displayed by the loadicator is an initial metacentric height derived from the KG calculated from entered weights and the KM read from the hydrostatic tables at the calculated displacement. It is a static figure valid for very small angles of heel in calm water.
A positive GM does not mean the ship will not capsize. It means that at very small angles of inclination in still water, there is a restoring moment. Whether that restoring moment persists through the range of angles actually encountered at sea depends on the GZ curve, not on GM alone.
Ships have been lost with adequate GM because the GZ curve had insufficient range, or because a large free surface effect collapsed the effective GM at the moment of flooding, or because the wind heel moment in survival conditions exceeded the righting arm. The loadicator’s GM readout is the beginning of a stability assessment, not its conclusion.
The minimum GM figure in the stability booklet is a regulatory floor, not a target. Operating at the minimum complies with the rules. It does not provide the margin that actual seakeeping in loaded North Atlantic winter conditions may demand.
6. Righting arm curves and the conditions they assume
The GZ curve generated by the loadicator is derived from the ship’s cross-curves of stability, which themselves are computed for an intact, unloaded-by-seas, upright hull form in calm water. Several critical assumptions are embedded in that derivation and are invisible in the final curve.
The curve assumes watertight integrity is maintained throughout the heel range shown. If any weathertight opening — a hatch cover with a degraded gasket, an air pipe without a functioning float valve, a ventilator not properly closed — becomes immersed before the angle of vanishing stability, the actual capsizing point is earlier than the curve suggests. The loadicator cannot account for the condition of individual weathertight fittings. That knowledge lives with the chief officer and the bosun.
The curve also assumes a homogeneous, non-shifting cargo. A bulk cargo that liquefies, a container stack that collapses, or a log deck that shifts in a roll changes the weight distribution dynamically. The loadicator calculated a GZ curve for the condition as declared. The condition as experienced no longer matches the declaration.
Cargo shift is not a failure mode the loadicator was designed to track. It is a failure mode that kills ships.
7. Cross-checking: draught, trim, and the physical ship
The only reliable cross-check for loadicator output is the physical ship. After loading, before departure, the actual draughts must be read from the marks and compared with the loadicator’s predicted draughts. If they do not agree within the tolerance specified in the stability booklet — typically around 50mm for deep draughts — the input data contains an error.
Trim is particularly diagnostic. An unexpected trim by the head or stern, where the loadicator predicts even keel, points directly to a weight distribution error: a tank not as declared, a hold density figure that is wrong, a ballast exchange not entered. The draught survey and the loadicator output should tell the same story. When they do not, the physical draughts are correct and the loadicator input is wrong.
List is a related diagnostic. A ship sitting with a persistent list after completing loading, with the loadicator showing zero TCG, has an undeclared asymmetry. It is possible to enter a counter-list ballast condition into the loadicator that masks the underlying asymmetry and shows a false zero on screen. PSC surveyors are aware of this practice.
The draught marks do not lie. The loadicator displays what it was told.
After a draught survey in port, any discrepancy between the surveyed displacement and the loadicator’s calculated displacement must be resolved before departure. A systematic discrepancy that appears across consecutive voyages usually points to an outdated lightship constant or a density assumption that is consistently wrong for the trading area.
Running the loadicator on a closing check — re-entering the final condition after all ballast exchanges and last-minute cargo adjustments — is not optional discipline. It is the only mechanism available to catch the effect of changes made in the final hours before sailing, when the pressure to meet pilot time is highest and attention to stability records is lowest.
8. Closing Reality
The loadicator is a calculation tool with a defined scope. Within that scope it is accurate. Outside that scope it is silent, and silence in an instrument does not mean the hazard is absent.
A green readout confirms that the entered loading condition, if the entered data is correct, falls within permissible still-water limits at the moment of entry. It does not confirm that the data is correct. It does not confirm that the condition has not changed since entry. It does not account for the loads the sea will impose, the free surfaces that will open when tanks are consumed, the shift that may occur in a beam sea, or the weathertight integrity of the hatch covers that keep the GZ curve valid.
Approval status, input discipline, and physical cross-checking with actual draughts are not procedural overhead. They are the mechanisms that keep the loadicator’s output connected to reality.
The officer who treats a loadicator output as a stability certificate has misunderstood both the instrument and the sea.