Sodium-Ion Battery for Electric Boats and Ferries
sodium-ion battery packs have crossed the line from laboratory curiosity to a genuine engineering option for short-route electric passenger vessels. Naval architects and ferry operators keep asking me the same question: can we stop specifying lithium iron phosphate and use a chemistry that does not depend on lithium, cobalt, or copper? The honest answer is narrower than the marketing material suggests, and it turns on three things: the route profile, the berth charging window, and the annual cycle count.
What follows is the sizing math for a realistic small ferry, the DC architecture that falls out of it, the cold-weather and condensation behaviour that decides whether a hull installation survives its first winter, and the class society evidence required before a passenger vessel is approved. I also cover the duty profiles where a sodium-ion battery is still the wrong choice for electric boats and ferries.
Why Sodium-Ion Fits Short-Route Electric Ferries
A sodium-ion cell pairs a layered oxide cathode built from iron, manganese, and sodium with a hard carbon anode. There is no lithium, cobalt, or nickel in the cell, and both current collectors are aluminium foil, which removes copper foil and the copper-to-aluminium ultrasonic weld from the bill of materials and eliminates a known corrosion failure mode.
A fully discharged cell can also be stored and shipped at zero volts without damage, because that is its least energetic state. Seasonal vessels can therefore be laid up with packs discharged, which simplifies winter storage, insurance, and UN 38.3 transport paperwork. A lithium battery cannot be treated that way without permanent capacity loss.
Cells available in volume today sit at 3.1 V nominal, operate between roughly 1.5 V and 3.9 V, and deliver 90 to 160 Wh/kg with 250 to 400 Wh/L. Cycle life is quoted at 3000 to 6000 cycles to 80% of initial capacity at 0.5C and 25 degrees Celsius. Cold performance is the standout: 85 to 92% of room-temperature capacity remains at minus 20 degrees Celsius against 65 to 80% for lithium iron phosphate. The trade is energy density, which on a 19-metre hull is absorbed without difficulty and on a 7-metre RIB is not.
Route Energy Budget: Sizing the Pack from the Duty Cycle
Every marine pack sizing exercise starts with a route energy budget, not a cell datasheet: modelled shaft power, hotel load, a weather and fouling margin, and a diversion reserve, all before dividing by the efficiency chain.
Worked example: 19 m, 49 passengers, 40-minute crossing
- Propulsion: 45 kW shaft power at 8 knots for 0.67 h = 30.2 kWh
- Hotel load: 4 kW average for 0.67 h = 2.7 kWh
- Weather, fouling, and windage margin at 15% = 4.9 kWh
- Diversion and holding reserve, not normally consumed = 10 kWh
- Usable energy required = 47.8 kWh
Nameplate capacity is the usable figure divided by the whole efficiency chain, and the chain is where first-pass designs go wrong. I use 0.92 for the inverter and motor controller, 0.90 for depth of discharge, and 0.85 for end-of-life retention, a combined factor of 0.704, so the pack is specified at 68 kWh. I would round to 72 kWh in three 24 kWh strings, so one string can be isolated for service. The resulting duty rate is comfortable: 0.63C continuous, peaking at 1.25C for 30 seconds during maneuvering, well inside the pulse capability of a hard carbon cell.
Weight, volume and thermal load
At 120 Wh/kg cell level, 72 kWh of cells weighs 600 kg. Enclosure, busbars, BMS, contactors, thermal hardware, and mounting add 20 to 25%, so the installed module lands near 750 kg. The same capacity in lithium iron phosphate at 155 Wh/kg would be 465 kg of cells, about 580 kg installed, a sodium-ion penalty of roughly 180 kg, under one percent of displacement, plus about 50 litres of volume. Placement matters more than the absolute figure, so I confirm the location against the intact stability calculation before fabrication.
Because the pack runs at 0.6C rather than 2C, heat generation is 2 to 3% of throughput, about 1.2 kW to reject. Air cooling at 350 to 400 cubic metres per hour with a 10 kelvin rise handles it comfortably, so no liquid cooling loop or coolant-leak risk is required. The compartment must still stay below 45 degrees Celsius, because calendar ageing accelerates above that figure.
DC Architecture, Voltage Class and Cable Sizing
Voltage class follows current, not habit. A 96-cell series string gives 297.6 V nominal, 374 V at a 3.9 V charge cut-off, and 240 V at a practical 2.5 V discharge floor. At that level 45 kW draws 150 A through a 50 square millimetre marine cable with acceptable drop; on a 96 V bus the same load would draw 470 A. Above roughly 100 kW of propulsion, a 400 to 700 V bus becomes the sensible choice.
Passenger vessels use an insulated, ungrounded distribution system, so the battery must float with respect to hull, an insulation monitoring device must be fitted, and positive and negative runs segregated to meet the class rule leakage thresholds.
Two BMS details deserve attention. The open-circuit voltage curve of a hard carbon anode is flatter through the middle of the range than a graphite-based cell, making voltage-based state of charge estimation unreliable between roughly 30% and 70%. Coulomb counting with periodic rest-state correction is mandatory, validated across the full temperature window rather than at 25 degrees only. Because a ferry uses a narrow state of charge window, small errors become large errors in remaining range, so I specify redundant current measurement plus a twice-yearly deep discharge to the low voltage threshold to reset the estimator. Contactor control must be independent of the propulsion controller, with a hard-wired interlock and a pre-charge resistor limiting inrush.
Cold Water, Cold Air and Condensation
A pack mounted low in the hull sits close to seawater temperature year round, typically 4 to 15 degrees Celsius, which is friendly to cell ageing but hostile to charging. Sodium-ion tolerates charging below freezing far better than lithium iron phosphate, but tolerance is not permission. Below zero I limit charge current to 0.1 to 0.2C, and below minus 10 degrees I stop charging unless the pack is warmed, because sodium plating on the hard carbon surface is capacity loss that never reverses. For an unheated winter berth, a 150 to 300 W shore-powered heater brings the pack to 5 to 10 degrees in 30 to 45 minutes.
Condensation, not cold, is the failure mode that actually kills hull-mounted packs. A sealed IP65 enclosure with a breather valve handles pressure swings from diurnal temperature change while keeping salt-laden moisture out. Inside, I specify conformal coating on the BMS boards, nickel-plated busbars, and a thermostatically controlled heater that keeps the interior 3 to 5 kelvin above ambient so moisture condenses on the enclosure wall rather than on the electronics. Bilge and salt-spray enclosures should be 316 stainless or marine-grade aluminium finished to C5-M under ISO 12944, and cells must be restrained against hull slamming with 0.5 to 2 mm compression pads, flexible busbars, and vibration isolators rated for the drivetrain frequency band.
Safety, Class Society and Flag State Approval
No single marine standard covers a sodium-ion pack end to end, so approval is built from a stack of documents. Cell and module evidence comes from IEC 62619 and, for transport, UN 38.3; shipboard electrical installation follows IEC 60092-101; ISO 16315 applies to small craft, and surveyors are reading ABYC E-13 across to sodium-ion. Passenger vessels also fall under SOLAS fire safety requirements, class society rules from DNV, ABS, Lloyd’s Register, or Bureau Veritas, and flag state approval.
On fire behaviour, sodium-ion has a real but bounded advantage. Thermal runaway onset is typically 20 to 50 kelvin higher than a nickel manganese cobalt cell and comparable to lithium iron phosphate, and total heat release per cell is lower. No cobalt means one less toxic combustion product in the hazard analysis. None of that makes the pack benign: the electrolyte is flammable and off-gas contains hydrogen and carbon monoxide, so the space needs detection and ventilation sized against a 4% hydrogen lower flammable limit. Water mist and aerosol suppression are both accepted in battery compartments on passenger vessels, either paired with A-60 boundary insulation where the space adjoins an accommodation or escape route.
Because there is no marine equivalent of UL 9540A for sodium-ion yet, class societies assess cell abuse data and a pack-level propagation test instead. Expect to supply the IEC 62619 certificate, a UN 38.3 summary for the pack configuration, abuse and propagation reports, a BMS functional description with failure mode analysis, insulation monitoring records, a ventilation and hazardous area assessment, and a hazard identification study.
Berth Charging: The Constraint That Decides the Route
Charging infrastructure, not pack chemistry, usually determines whether a route is electrifiable. A 72 kWh pack charged at 0.5C absorbs 36 kW, and a 40-minute turnaround leaves only a fraction of that time available once passengers are exchanged and the vessel secured. Recovering 38 kWh in 25 minutes of real charging time takes 91 kW, or 1.27C, and sodium-ion accepts roughly 1C continuously and 2C where the cell is optimised for it, derating above 40 degrees Celsius. A 150 kW charger therefore buys nothing useful.
An AC connection at 63 A three-phase and 400 V delivers 43 kW and needs nearly two hours, which suits a ferry with a two-hour layover and avoids a DC charger entirely. For a fast turnaround, DC charging at 90 to 150 kW requires a berth service upgrade, a galvanic isolation transformer, and load management so charging does not collide with other harbour loads. The cheapest way to cut both pack size and charging power is to charge at both ends of the route: splitting 72 kWh across two berths lets a 45 kWh pack do the same work with a smaller cable and a lighter vessel.
Cycle Life and Cost per Delivered kWh
This is where the chemistry conversation gets uncomfortable and where the real decision is made. A cell quoted at 3000 to 6000 cycles to 80% capacity is competitive in stationary storage, where a pack sees 250 to 400 equivalent full cycles a year. A busy ferry is another matter. Take the worked example: 12 crossings a day, 350 operating days, each drawing 38 kWh from a 72 kWh pack. That is 53% depth of discharge per crossing, or roughly 2200 equivalent full cycles a year, so at 4000 cycles the pack reaches end of life in under two years. Lithium iron phosphate at 6000 to 8000 cycles buys three to four years on the same duty, so the cheaper chemistry loses on lifetime cost per delivered kilowatt hour despite a better purchase price.
Now change the duty profile. A lake ferry running four round trips a day, or a seasonal vessel operating 120 days a year, sees 300 to 600 equivalent full cycles annually and the same pack lasts eight to twelve years. There sodium-ion wins clearly: lower capital cost, no dependence on lithium or cobalt pricing, better cold-weather capacity, and zero-volt layup between seasons. The decision rule is worth writing down before any cell is ordered. Above roughly 1500 equivalent full cycles a year, specify lithium iron phosphate. Below 800, sodium-ion is usually the better commercial choice. In between, run the numbers including the cost of capital.
Calendar ageing matters as much as cycling on a vessel that sits at the dock overnight. Storing at 30 to 70% state of charge and keeping the compartment below 35 degrees costs nothing and buys years of service life.
What Sodium-Ion Still Cannot Do
- Long crossings. Above roughly 60 minutes at 10 knots, energy density becomes the binding constraint, paid for in displacement.
- Charging above about 2C. A route needing a 20-minute full recharge must be built around a different chemistry.
- Ultra-high cycle fleets. Operations above 1500 to 2000 equivalent full cycles a year should stay with lithium iron phosphate.
- Weight-critical craft. Foiling and planing hulls are better served by a lighter cell.
If none of those exclusions apply, the case is strong. Any pack in this class is ultimately a custom battery solution, sized around a specific route rather than picked from a catalogue.
Frequently Asked Questions
Can sodium-ion batteries be charged below freezing on a boat?
Yes, at reduced current, which is a meaningful operational advantage over lithium iron phosphate. I limit charge to 0.1 to 0.2C between zero and minus 10 degrees Celsius, and stop below minus 10 degrees unless the enclosure is warmed first. A 150 to 300 W shore-powered heater brings a hull-mounted pack up to temperature in 30 to 45 minutes and prevents permanent capacity loss from sodium plating.
How much heavier is a sodium-ion pack than lithium for the same ferry route?
Using the 72 kWh example above, roughly 180 kg installed, under one percent of displacement on a 19-metre vessel. Energy density of 120 Wh/kg for sodium-ion against 155 Wh/kg for lithium iron phosphate drives that. A 22-tonne hull absorbs it through trim adjustment; a weight-critical foiling craft does not.
Do sodium-ion batteries need a different BMS than lithium-ion?
The hardware architecture is similar but the state of charge algorithm is not. The flat open-circuit voltage plateau of a hard carbon anode between roughly 30% and 70% state of charge makes voltage-based estimation unreliable, so coulomb counting with periodic rest-state correction is required. Charge limits must also encode the reduced-current charging permitted below freezing, which generic lithium profiles do not.
Are sodium-ion batteries allowed on passenger vessels under SOLAS?
Yes, through the same approval route as other lithium-ion chemistries, provided the class society and flag state accept the evidence package: IEC 62619 cell certification, a UN 38.3 transport summary, abuse testing with a pack propagation test, an insulation monitoring scheme, a ventilation calculation against a 4% hydrogen lower flammable limit, and fire detection and suppression for the battery space.
What is the realistic cycle life for a ferry duty cycle?
Roughly two years on a heavily used commuter route and eight to twelve years on a seasonal route, for the same 3000 to 6000 cycle cell. A ferry running 12 crossings a day at 53% depth of discharge accumulates about 2200 equivalent full cycles a year, while a seasonal lake ferry running four round trips a day accumulates 300 to 600. Cycle count governs, not calendar age.
Can I retrofit sodium-ion into an existing electric boat?
Usually yes, and in three respects the retrofit is simpler than a lithium installation: no copper current collectors, no need to heat the pack before charging in mild cold, and zero-volt layup. What must be revisited is the charger profile, the BMS state of charge algorithm, the insulation monitoring thresholds, and the enclosure ventilation and detection design. Do not reuse a lithium charge profile without confirming the per-cell cut-offs, because the 3.1 V nominal cell changes string voltage.
Further Reading
References
