Knowledge / Ship systems

Watertight boundaries: from initial damage to progressive flooding

Explore flooding paths between compartments, differential head, intermediate stages and closure evidence through a worked example.

On this page

Watertight subdivision works when it can separate the initial ingress from the rest of the ship. Progressive flooding is the spread of water beyond the initially damaged space through openings, pipes, ducts or failed boundaries. The size of the shell breach therefore matters alongside the state of the connections inside the ship.

Drawing a compartment does not close its water paths

A bulkhead, deck, door, hatch and every penetration belong to the same boundary. An open pipe through a strong bulkhead can connect it to another space even when the pipe is relatively small. A cable-transit gap and a ventilation opening above the waterline have different initial conditions; heel or sinkage may subsequently immerse the second opening.

IMO’s explanation describes damage control plans and booklets as information about subdivision and the equipment that preserves it, helping officers limit progressive flooding. Possessing a plan does not establish that every closure aboard is in the position assumed on that plan. The document is a reference against which the actual condition must be understood.

Watertight, weathertight and fire boundaries

A watertight closure provides its sealing function under the water pressure for which it is designed. A weathertight closure does not establish that the same immersion and pressure conditions are acceptable. A fire-rated penetration is likewise not automatically a watertight penetration. One closure may serve several functions, but the evidence for each function has its own scope.

MCA MGN 689 addresses harmonisation of cargo-ship door requirements across SOLAS, Load Lines, MARPOL and the relevant cargo codes. It does not produce one operating rule for every door on every vessel. Ship type, construction date, the door’s location and the approved operating arrangement are needed to identify the applicable requirement.

Read the flooding network through time

In an illustrative network, compartment A is breached to the sea, a door connects A to B, and a ventilation duct connects B to C. Closing the door may confine flooding to A. If it remains open, B floods; the additional trim may immerse the connection to C. An opening initially above water can become critical later. Connection elevation and closure state are properties of a network that changes with time.

MCA MSIS42, revision May 2023 treats intermediate flooding stages, equalisation and progressive paths explicitly. The final equilibrium condition therefore does not describe the whole process. A ship that appears less heeled at the end may pass through an unacceptable intermediate heel, opening immersion or loss of an essential service.

A bounded model of flow through an opening

For a small opening with negligible approach velocity, a simplified relationship is Q = C_d A √(2g Δh). Q is volume flow, C_d the dimensionless discharge coefficient, A the opening area and Δh the water-head difference across it. The US Bureau of Reclamation explains its use for standard water-measurement orifices; it does not supply a coefficient for an arbitrary ship breach.

Assume an opening of 0.020 m², C_d = 0.60, a constant head difference of 2.0 m and g = 9.81 m/s². Then Q = 0.60 × 0.020 × √(2 × 9.81 × 2.0) ≈ 0.0752 m³/s, or 271 m³/h. If those fixed conditions lasted five minutes, about 22.6 m³ would enter. These inputs are illustrative, not a flooding prediction for a real ship.

Even a drainage path actually delivering 100 m³/h would remain below that initial inflow. The initial net accumulation would be about 171 m³/h. Pump nameplate capacity alone is insufficient: delivery head, suction, electrical supply, strainers and the discharge route must all remain usable in the damage condition being considered.

Why the flow does not stay constant

As the internal water level rises, the differential head may fall; as the ship sinks, the external level may rise relative to the opening. Trapped air, waves, flow reversal and breach shape can alter the simple model. USBR’s coefficient discussion emphasises the connection between coefficients, geometry and measurement conditions. The illustrative value 0.60 is not an acceptance criterion.

Converting a compartment volume to a flooding time also requires care. Structure, cargo and equipment occupy space; permeability represents the fraction available for floodwater. Mixing added-weight and lost-buoyancy conventions can count the same effect twice. An elementary discharge calculation cannot replace a damaged-stability or structural-strength assessment, and the initial-rate calculation cannot establish time to capsize.

From a close command to a physical boundary

A closed indication on a screen reports a particular sensor state. Seating of the door, operation of its securing mechanism, seal integrity and performance under pressure need different evidence. An interpretation based only on a green symbol is weak unless switch adjustment, behaviour after loss of power and the local control state are understood.

A temporary hose, an open manhole or an unfinished cable transit during maintenance can change the effective subdivision. Restoration should record that the penetration has regained its required function with suitable materials and the necessary checks. Human access also matters: maintaining a damage boundary must be coordinated with safe passage, without exposing personnel to movement of a powered closure.

Which results must be read together?

Flooding assessment asks more than whether the ship remains afloat. Residual stability, immersion of openings, differential water loads on bulkheads, power and communication loss, and escape routes belong to the same scenario. Moving water to an adjacent compartment may reduce heel while increasing the total flooded volume; equalisation must be considered within the vessel’s approved arrangement.

A useful engineering explanation separates the initiating damage, active connections, assumed closure positions, changing water levels and available drainage. Evaluating both plausible states of an unknown door is more informative than silently assuming it closed. The result should show which failed boundary exposes which additional volume, and which missing observation could materially change that conclusion.

Sources