Designing a lab-scale ElectroSAnMBR for ship bilge water

ElectroSAnMBR is a European Union Horizon 2020 research project developing a submerged anaerobic electrochemical membrane bioreactor for treating ship bilge water. It combines electrolysis, anaerobic digestion and membrane separation to improve the removal of oils, chemicals and organic pollutants. The project is a research platform, not a turnkey shipboard treatment service.

A useful lab-scale design must reproduce the important mechanisms, not simply shrink a full-scale vessel. The focus is reactor geometry and electrode configuration: the physical features that govern mixing, mass transfer, electrical behaviour and fouling. Australian shipping conditions add useful constraints, from varied bilge chemistry in major ports to tight expectations for oily waste handling.

Why geometry matters for ship bilge treatment

Geometry sets the operating envelope. Vessel proportions, baffles, inlet and outlet positions, membrane orientation and electrode spacing determine whether bilge reaches the active zones or bypasses them. A submerged reactor must also cope with rising gas, settling solids and recirculation, so a geometry that suits an aerated membrane tank may behave differently under anaerobic electrolysis. Background on microbial electrolysis and membranes is useful when evaluating these coupled effects.

Bilge water is not a uniform feed. Free oil, emulsified oil, fuel residues, cleaning chemicals and salt can change together as a ship changes cargo or operations. Solids may shield membrane surfaces, while conductive dissolved salts reduce resistance in the electrical circuit. An oil-rich fraction should be separated before the membrane stage, but pretreatment must not hide the fouling problem by removing the hardest contaminants from every trial.

Australian ports such as Port Hedland, Newcastle and Fremantle receive vessels with very different waste streams, from bulk-carrier residues to mixed coastal-service washings. Marine crews and ship-repair staff use terms such as “bilge”, “oily sludge” and “separator waste” for distinct streams, so samples need clear labels. Discharge controls under Australia’s implementation of MARPOL also mean that laboratory removal results belong beside oil-separation performance; they are not discharge permission.

Choosing a compact geometry for the test rig

A sensible starting point is a wide, shallow, electrically nonconductive tank with rounded corners and a sloped floor. The plan area accommodates a flat-sheet membrane cassette and parallel electrodes, while the low liquid depth simplifies sampling and electrical connections. Rounded geometry reduces sediment traps, and a removable oil-rich drain allows contamination to be collected without dismantling the reactor.

The internal arrangement matters as much as the outer shape. Distribute feed along the lower membrane edge, return it through a baffled recirculation loop and avoid short-circuiting directly from inlet to outlet. Low-speed crossflow can provide shear and mixing without the oxygen penalty of vigorous aeration. Keep the inlet, outlet, gas vent and sampling ports apart so that each measurement represents the reactor rather than a convenient dead zone.

Scale-down should preserve selected hydraulic and electrochemical ratios, not every dimension. Record liquid height, working volume, membrane area, recirculation rate, superficial flow velocity, transmembrane pressure, current density and gas hold-up. A compact rig can then reveal which variables matter and which merely reflect vessel size. Temperature control and cleanable seals are especially important in Australian laboratories, where warm rooms can accelerate biological activity and distort comparisons.

Configuring electrodes around the membrane zone

Use removable flat electrodes held parallel to the membrane faces on insulating mounts. Alternate anode and cathode positions in a modular frame, fix the gaps and prevent direct electrical contact. This arrangement allows the same reactor to test different spacing, polarity and active-area ratios without rebuilding the vessel. If hollow-fibre membranes are used, revisit the layout because their geometry needs different distributors and much lower feed velocities.

Material choice sets the baseline. Graphite plates are a straightforward laboratory option, while carbon or coated-metal surfaces can increase area and microbial attachment. Anodes also need a contaminant-specific durability test. Record full-cell voltage and current continuously, normalise current by actual active area, and use a protected reference electrode to distinguish electrode potential from ohmic losses in the bulk liquid.

Keep the intended biology visible in the design. Cathodic hydrogen may support hydrogenotrophic methanogens, whereas anodic oxygen can inhibit strict anaerobes if it reaches the biological zone. Monitor dissolved oxygen near the anode and in the bulk, and test whether mixing dilutes or concentrates it. This is especially relevant on Australian routes from Darwin towards Southeast Asia and across Bass Strait, where temperature changes also alter oil viscosity and reaction rates.

Treat the gas pathway as part of the reactor. Cathodic hydrogen, anodic oxygen and biologically produced methane may coexist, so use a vented gas-analysis header, pressure relief and suitable sensors rather than allowing gas to accumulate beneath the cassette. Electrical isolation, current limits and a documented shutdown procedure complete the laboratory safety case.

Monitoring performance under realistic bilge loads

A credible test programme measures chemistry, biology, hydrodynamics and electricity together. Track oil and grease, chemical oxygen demand, total organic carbon, pH, conductivity, dissolved oxygen, methane, membrane flux, transmembrane pressure, current, voltage and specific energy use. Samples from the influent, membrane zones and final outlet reveal where removal occurs and where fouling or electrical losses develop.

Use real bilge samples alongside a controlled surrogate matrix based on fuel oil, lubricating oil and representative cleaning chemicals. Vary salinity, temperature, surfactant content and load in controlled steps, with replicated membrane-cleaning cycles between trials. A power-on reactor should be compared with membrane-only and electrolysis-only controls so the contribution of each process can be identified.

Port operations make the commercial context clearer. Around Henderson near Fremantle, Newcastle and Darwin, separated oily waste is collected for specialist shore-side treatment rather than handled as an unrestricted discharge. A laboratory treatment train should therefore match an actual waste stream and produce auditable oil-separation data. Equipment intended for shipboard use would also need to meet vessel-class, electrical and space requirements not settled by a bench experiment.

Researchers should resolve equipment assumptions with the research contacts at UYUT before releasing fabrication drawings, and align the test plan with the project’s experimental methods.

A practical build-and-test roadmap

Four layouts offer different experimental compromises:

Geometry Main advantage Main limitation Best experimental role
Wide baffled tank Easy access and parallel electrode placement Uneven flow near walls Baseline membrane and electrode trials
Tall narrow column Small footprint and longer vertical flow path Stagnant zones and higher head loss Gas hold-up and axial mixing studies
Cascaded narrow cells Independent flow and sampling More pumps, tubing and sensors Replication and scale-down checks
Staged baffled vessel Adjustable residence time and staged sampling More oil and solids accumulation points Fouling and shock-load tests

Choose the wide baffled tank as the baseline only if membrane flux, current distribution and fouling can be held within a repeatable range. The narrow column and cascaded cells are useful for hypothesis testing, not merely compact alternatives. A staged layout adds residence-time control, but every extra compartment also adds a potential fouling point.

The fabrication package should define the scale ratios, electrode area and gap, membrane area, ports, drains, sensor positions and safe operating envelope. This prevents a visually convincing vessel from becoming an instrument whose data cannot be compared with another run.

Design choices worth locking in

The next step is to turn the selected layout into a dimensioned schematic, bill of materials and staged test schedule, then submit it for laboratory and safety review. That gives the ElectroSAnMBR work packages a testable design rather than another untested concept.