Knowledge / Ship systems

Bilge and oily-water management: separation, monitoring and barriers

An engineering view of bilge-source control, representative sampling, protective response and traceable oily-water disposal.

On this page

Machinery-space bilge management is a chain of containment, treatment, measurement and authorized disposal. Its weakest link may be a leaking machine, a misleading sample or an unavailable shore facility rather than the separator itself. This article explains that chain for engineering education. It does not provide a discharge checklist, replace the approved piping arrangement or authorize operation of an overboard connection. Actual handling must follow MARPOL as applicable, local restrictions, the ship’s safety-management procedures and equipment instructions.

Separate permission from equipment capability

The presence of a 15 ppm oily-water separator does not mean discharge is permitted everywhere whenever its display reads below that value. MARPOL Annex I distinguishes equipment requirements from discharge conditions; ship size, operating area and the nature of the mixture matter. AMSA’s discharge-standards summary illustrates the need to consider the vessel’s operation and location, including additional polar and local restrictions. Read the applicable legal text for the specific case.

This discussion concerns machinery-space bilges, not a tanker’s cargo-tank slops. Those waste streams have different control arrangements and must not be casually combined. A sensible planning decision can be to retain oily water for a suitable reception facility even when treatment equipment exists. Such a decision requires enough safe storage, a realistic quantity estimate and arrangements ashore. It should be made before the tank level becomes an operational emergency.

Reduce the burden before treatment

A bilge tank collects the consequences of many activities. Seal leakage, fuel and lubricating-oil drains, wash water and maintenance residues can arrive at different rates and with different compositions. Investigating the origin of each stream often yields a larger improvement than adjusting the downstream separator. A persistent increase in water volume may indicate a mechanical defect or poor segregation; it is not simply a demand for longer treatment hours.

Oil droplets can be dispersed finely in water, and cleaning chemicals can stabilize mixtures that are difficult to separate. Solids can foul treatment stages and sampling lines. Select cleaning products and disposal routes through the ship’s approved procedures, considering compatibility with treatment equipment. Do not assume that something which makes a bilge look cleaner makes its contents easier to treat. Keep chemical residues out of a process that was not designed or approved to handle them.

Understand the treatment train

Gravity separation relies on differences in density and the opportunity for droplets to move relative to water. Coalescing elements encourage smaller droplets to join, while later stages may use other physical treatment mechanisms. The actual technology and consumables depend on the approved unit. A pressure indication or a running motor therefore cannot, by itself, establish successful oil removal. Compare the observed condition with the maker’s documented operating range.

Temperature, throughput and feed condition interact. Changing one variable to improve separation can exceed another equipment limit, damage consumables or create a different hazard. Maintenance should restore the approved condition, not create an undocumented alternative process. The relevant equipment specification is IMO resolution MEPC.107(49), with applicability depending on the installation and approval history. It is not a guarantee that every possible bilge mixture can be processed successfully.

The sample is part of the protective function

The oil-content monitor assesses the water presented to it. If that sample is unrepresentative, the displayed value cannot establish the quality of the main stream. Sampling-path integrity must therefore be considered alongside sensor calibration. A blocked line, unsuitable flow or a cleaning arrangement left in the wrong state can sever the connection between the displayed number and the water being handled.

AMSA’s 2025/06 notice on testing and inspection emphasizes representative sampling, fail-safe operation and whole-system response. Its inspection interpretation is Australian guidance supporting the IMO framework. A test should demonstrate the intended protective outcome using the approved procedure, including the final stopping or diversion function where fitted. Proving that an alarm lamp works is a narrower result and should be recorded as such.

An original response-time example

Assume, solely for a teaching calculation, that a sample path holds 0.6 litres and carries a steady 3 litres per minute. Ideal plug-flow transport time would be 12 seconds before considering the instrument or final valve. If fouling reduces sample flow to 1 litre per minute, the corresponding transport time becomes 36 seconds. The calculation shows why a small line can dominate the response even when the electronic monitor itself is fast.

Real samples mix, lines have different geometries and equipment requirements constrain acceptable arrangements. These numbers are not approved design values or a compliance test. The lesson is to examine the complete sequence from an effluent change to prevented discharge. An installation cannot be credited with adequate response solely because a component certificate lists a short response time. Any modification of the sample path requires the appropriate technical and approval review.

Use a transparent quantity balance

A useful management model tracks starting quantity, incoming bilge water, transfers, treatment returns, shore delivery and closing quantity. Each movement needs a source and destination; recirculation must not be mistaken for final disposal. Soundings and flow indications have uncertainty, so the aim is a defensible reconciliation rather than false precision. Unexpected differences are a reason to investigate, not a reason to alter recorded figures until the totals look tidy.

For example, assume 8 m³ is initially retained, another 5 m³ is collected and 4 m³ is delivered to a reception facility, with no other movement. The expected remainder is 9 m³. A measured 10.2 m³ may reflect an undocumented inflow, a tank-calibration issue or a measurement error. The arithmetic identifies a question; it does not identify wrongdoing or prove a particular fault. Preserve the original observations while establishing what explains the discrepancy.

Plan maintenance and waste capacity together

An unavailable separator changes the storage forecast. If a ship generates an assumed 1.5 m³ of bilge water per day and has 6 m³ of usable spare capacity, the simple planning horizon is four days before allowing for margin or variable inflow. Usable capacity is not the tank’s total geometric volume; operational limits, existing contents and contingency allowance reduce it. A rising leakage rate can shorten the horizon rapidly.

Maintenance planning should therefore include reception options, spare consumables, competent support and a safe test arrangement. Equipment can be repaired while the waste-management problem remains unresolved. Record that distinction in the handover: treatment readiness, available storage and disposal arrangements are separate statuses. A shore receipt should identify what was actually received and be reconciled with shipboard quantities; it does not erase the need for an accurate operational history.

Diagnose without confusing a symptom with its cause

An elevated monitor reading can result from actual oil carryover, sample contamination or an instrument problem. These possibilities call for different corrective actions. Begin with the safe condition defined by the approved procedure and compare independent evidence: feed history, recent cleaning, treatment-stage condition, sample-path condition and instrument service status. Replacing the sensor first may remove neither an unstable feed nor a damaged treatment element. Conversely, changing treatment settings cannot correct an unrepresentative sample.

A practical fault review separates what is observed from what is inferred. “The reading increased after tank changeover” is an observation; “the new tank contains an emulsion” is a hypothesis requiring evidence. Preserve both the event and the uncertainty. Trend recurrence across tank sources and maintenance dates, while avoiding uncontrolled experiments that could lead to discharge. Diagnosis should improve understanding without removing the protective boundary around an uncertain mixture.

Make the maintenance evidence usable at handover

A completed work order is more informative when it names the restored function and the test used to demonstrate it. “Monitor cleaned” describes an activity. A record explaining that the approved test confirmed measurement and the intended protective response describes evidence. Include unresolved limitations, equipment configuration and the person responsible for the next action. Do not label the entire installation ready if the separator was serviced but the final protective function remains untested.

The same discipline applies after a long period of retention. Contents can change through new inflows or maintenance activity, so an earlier successful treatment run cannot establish the suitability of today’s feed. Review storage history before resuming operation. Good assurance makes the chain understandable to the next watch and to an independent reviewer, while retaining the difference between observed performance, assumptions and legal permission.

Recognize the common failures of assurance

Common mistakes include treating the ppm display as a discharge permit, testing only the monitor, omitting returned liquid from the quantity balance and accepting recurrent alarms as normal behaviour. Equally weak is assuming that a clean-looking sample proves absence of oil. Do not defeat a protective device, dilute a sample to create an acceptable indication or improvise an unapproved piping route. Such actions destroy the very evidence the system is intended to provide.

A sound review asks whether each barrier has a clear purpose and observable evidence: sources controlled, storage adequate, treatment maintained, sample representative, protective action effective and records coherent. IMO’s machinery-space bilge-equipment page identifies the governing equipment-resolution family and approved-equipment information. Use that reference together with the actual ship documents; a generic educational article cannot determine the vessel’s permissible discharge conditions.

Sources