Lithium Battery for Electric Outboard and Small Craft

Most electric outboard conversations start with a fuel bill. A centre console burning 60 litres of petrol on a day’s running is a boat that gets used less than the owner wants. Then the owner finds a 48 V golf-cart style pack online and asks whether it will push a 6.5 m hull. It will not, and the reason is the shape of the load rather than the capacity. A trolling motor asks for 40 to 60 A of near-constant current. A propulsion outboard asks for 200 A at cruise and 500 A for eight seconds when you open the throttle against a head sea.

Cutaway of a lithium battery pack for an electric outboard propulsion system showing prismatic cells, BMS and high-voltage busbars

This is the sizing and integration approach we use at Horizon Power when a boatbuilder or refit yard asks for a lithium battery that has to sit in a bilge, survive salt water, and deliver rated power until the last 5% of charge.

Propulsion Loads Are Nothing Like a Trolling Motor

For a fixed-pitch propeller, shaft power rises roughly with the cube of shaft speed and thrust with the square. Adding 10% to cruise speed adds about 33% to power draw. That steep curve is exactly where a bow-mount pack falls apart, and it shows up in three places:

  • Continuous versus peak current. A 25 kW outboard on a 96 V bus draws about 260 A at full throttle. The pack has to hold that without cell tabs exceeding 45 to 50 °C. Marketplace peak ratings are 10-second numbers and say nothing about a 40-minute run against tide.
  • Voltage sag becomes a control problem. Controllers map throttle to torque. A pack with 8 mΩ of DC internal resistance drops 2 V at 250 A, the controller compensates by drawing more current, and the boat feels weak at the top of the throttle range.
  • Duty cycle is not flat. A displacement hull cruises at 4 to 6 kW then demands 20 kW in a squall. A planing hull needs a burst to get on plane and then settles to 40 to 60% of that — easier on the pack than most owners expect.

One more load catches people out: regeneration. A sail drive or an outboard left in the water under sail pushes 50 to 250 W back into the pack. The charge path must be active and temperature-gated. If the BMS blocks charging below 0 °C and the regen controller does not know, the energy has nowhere to go and bus voltage climbs.

Choosing the DC Bus Voltage

Voltage drives every other cost, so settle it before cell chemistry. The rule we use is to keep continuous current below roughly 300 A, and to cross the 60 V DC touch-safety threshold only when power demand genuinely forces it.

  • 48 V nominal (13S LFP, 54.7 V charge). Right for 3 to 10 kW drives, tenders, and displacement hulls under 6 m. Manageable cables, widely available connectors, and a bus below the shock-hazard threshold in most jurisdictions.
  • 96 V nominal (30S LFP, 109.5 V charge). The sweet spot for 15 to 50 kW. Current at 25 kW is 260 A instead of 520 A, which halves cable mass and relieves busbar temperature rise. Most serious conversion projects land here.
  • 400 V nominal (125S). For foiling craft and 100 kW+ drives. It brings IEC 60664-1 creepage rules, double insulation, interlock loops on every serviceable connector, and orange HV cabling. Do not go here unless mass forces it.

Above 60 V, every connector must be touch-safe and rated to break under load. A pack that can deliver 500 A into a short is not something you disconnect with a spanner. We specify a service disconnect loop and a pre-charge circuit on every high-voltage pack, because inrush into a bare inverter input will weld a contactor closed.

Sizing the Pack: A Worked Range Calculation

Published electric boat range figures usually assume flat water, no wind, and a perfectly matched propeller. Here is the calculation we run with customers on a 6.5 m centre console. Target: 15 nautical miles at 12 knots. Measured propulsion power at 12 knots was 9 kW on a clean bottom, so 1.25 hours needs 11.25 kWh. House loads — chartplotter, sounder, VHF, nav lights, a small fridge — add 300 W for 1.5 hours, or 0.45 kWh. Delivered energy: 11.7 kWh.

Now three non-negotiable factors: depth of discharge 0.90, end-of-life capacity retention 0.80, and a 1.20 reserve for fouling, chop and a headwind home. Pack capacity = 11.7 × 1.20 / (0.90 × 0.80) = 19.5 kWh. At 96 V that is 203 Ah, which we build from 210 Ah prismatic cells in 30S for 20.2 kWh. At 135 Wh/kg pack level the pack weighs about 145 kg, so it goes low and near midships where the stringers allow.

Cell Chemistry and Format

  • LFP prismatic, 3.2 V nominal. 150 to 180 Wh/kg at cell level, 4,000 to 6,000 cycles to 80% at 0.5 C and 25 °C, and the flattest discharge curve available. Flat voltage means motor rpm does not fall as the pack empties. Prismatic cans also give a stable stack that does not need the compression fixtures pouch cells require, which matters when a hull slams off a wave.
  • NMC, 3.7 V nominal. 240 to 280 Wh/kg, useful when fighting for mass on a foiler or a weight-limited tender. The trade is thermal runaway onset near 150 °C against 200 to 250 °C for LFP, plus shorter cycle life, and the enclosure then needs cell-level containment and venting.
  • Sodium-ion. Retains 85 to 92% of capacity at −20 °C and has no lithium plating mechanism, so cold-water charging is far more forgiving. The penalty today is 25 to 40% more volume for the same energy, which is a fair trade in a bilge with space.

We do not use pouch stacks in a hull that pounds. Prismatic cans with welded busbars and anchored cell holders are the format that survives fifteen years of vibration.

Thermal Design Inside a Sealed Marine Enclosure

A boat is the worst thermal environment in transport: a sealed compartment, sun on the deck above, salt water below, no airflow. Cells want 15 to 35 °C. Above 45 °C, calendar ageing roughly doubles per additional 10 °C and cycle life falls from 4,000 to under 2,000.

Our standard approach is a welded aluminium enclosure that doubles as the heat spreader, with cells bonded to the base through a conductive gap pad rather than simply dropped in. At 1 C continuous, LFP produces 2 to 3% of throughput as heat, so a 20 kWh pack at 20 kW generates 400 to 600 W. Above 15 kW continuous we add a cold plate on a raw-water loop.

Sealing is the other half. A sealed enclosure breathes through its gaskets as it heats and cools, and a season of daily 30 K swings will pull in humid air and condense it. Every pack we build uses an ePTFE pressure-equalisation vent, conformal coating on the BMS, and tinned copper terminals — bare copper in salt air is a two-year failure. Stainless fasteners into aluminium need isolating washers, or the galvanic couple eats the enclosure around the bolt heads. IP67 is the floor for a bilge pack, IP68 where the compartment can flood, and anything going into service should be salt-fog tested to IEC 60068-2-52 and vibration tested to IEC 60068-2-64 on the assembled pack, not on bare cells.

BMS, Contactors, Shore Charging and Galvanic Isolation

The BMS in a propulsion pack is a power device as much as a monitoring device. It drives the main contactor, enforces the current limits the controller assumes, and is the last line of defence against runaway overcharge on a 96 V string.

  • Modular BMS. One board monitoring 30 series groups from a single connector is a harness nightmare in a wet boat. Slave boards per 6 to 8 cell group on an isolated CAN bus are far more serviceable.
  • Impedance-based state of health. Capacity lags. A 1 kHz impedance rise of 25 to 30% appears 300 to 500 cycles before the capacity knee, giving a season of warning instead of a mid-season surprise.
  • Pre-charge and sequencing. Pre-charge resistor, then main positive, then main negative, with the controller enabled only after a bus voltage check. Contactors drop on any cell over-voltage, under-voltage, over-temperature or isolation fault.
  • Charge gating. No charging below 0 °C unless heating pads are fitted and cell temperature — not ambient — is above freezing.

Shore charging is where marine installations most often go wrong. Use a marine-rated charger with ignition protection if it shares a space with petrol, and set a two-stage profile with no float stage: holding LFP at 100% state of charge indefinitely accelerates calendar ageing rather than maintaining anything. Route the DC charge path through the BMS so charge FET gating is preserved, and fit a galvanic isolator or isolation transformer on the AC inlet. Without one, the shore earth becomes a corrosion path to every underwater metal part, and the battery gets blamed for a corrosion problem it did not cause.

Standards, Certification and Insurance Reality

There is no single global certificate for an electric outboard battery, but surveyors and insurers ask about a predictable stack: IEC 62133-2 or UL 1642 at cell level; IEC 62619 or UL 1973 at pack level including propagation and drop tests, with UL 9540A data increasingly requested for larger installations; UN 38.3 plus a dangerous goods declaration for transport, shipping at 30% state of charge; ABYC E-13 for lithium installations on small craft and ISO 16315 for electric propulsion, covering compartment venting, over-current protection location, charge-source disconnection and low-temperature charging lockout. Commercial vessels add class society approval and a propulsion failure modes analysis, which is where cost and timeline jump.

Insurance is the practical gate. Hull insurers increasingly ask three questions: does the pack carry IEC 62619 or UL 1973, was the installation signed off by a qualified person, and is there a documented low-temperature charging interlock plus compartment gas detection. Having those three in a folder is worth more than another kilowatt-hour.

Commissioning and What Actually Fails

The first 50 cycles tell you whether the installation is sound: balanced string voltage at rest, insulation resistance above 100 MΩ at 500 V, and a documented 0.2 C capacity test. A first capacity test below 95% of label is almost always a harness error or a bad sense lead, not the cells.

Across the packs we have had returned from the field, failures cluster predictably:

  • Sense-lead damage from vibration — the most common single cause of a pack going out of balance and being condemned. Strain-relieve every lead at both ends.
  • Water ingress through unvented enclosures — almost always condensation, rarely immersion.
  • Terminal heating — a torque spec is not optional. Run a thermographic inspection at 1 C continuous and re-torque anything reading more than 15 K above ambient.
  • Owner habits — leaving the pack at 100% on a float charger all winter does more damage than 2,000 cycles ever will. Store at 40 to 50% state of charge, disconnect the main, and check every 90 days.

The economics still work. A 19.5 kWh pack is roughly USD 5,500 to 6,800 at pack level before the outboard and charger, against 30 to 45 litres of petrol for that same 15 nautical mile day. A summer of weekend use is 60 to 100 cycles against a 4,000-cycle claim, so the limiting factor will be the calendar, not the cycles.

Frequently Asked Questions

Can I use a standard 48 V lithium battery for an electric outboard?

Only for small drives. Up to about 10 kW, a 48 V pack with a 200 A continuous BMS and proper busbars works well. Above that the current demand pushes you into cable and contactor sizes that are impractical, and you should move to a 96 V bus.

How many kilowatt-hours do I need for a day of electric boating?

Work from measured power at your real cruise speed. As a working range: 1.2 to 1.5 kWh per nautical mile for a 6.5 m planing hull at 12 knots, and 2.5 to 4 kWh per nautical mile for a heavy displacement hull at 6 knots in current.

Is LFP or NMC better for a marine propulsion battery?

LFP for nearly everything, because prismatic LFP tolerates vibration, holds voltage flat under load, and has a much higher thermal runaway onset. NMC is justified when mass is the binding constraint, such as foiling craft, and then the enclosure must provide containment and venting.

Does a lithium battery pack need to be vented in a boat?

The enclosure needs pressure equalisation and the compartment needs ventilation. Use an ePTFE vent on the pack so it does not inhale humid air, and follow ABYC E-13 for compartment venting and gas detection. A fully sealed compartment with a lithium pack inside is an inspection finding.

How does cold weather affect an electric outboard battery?

LFP loses discharge capacity below freezing and must not be charged below 0 °C, because lithium plating permanently damages the anode. Sodium-ion keeps 85 to 92% of capacity at −20 °C and tolerates charging better, which is why cold-water operators ask for it.

Can my outboard regenerate and charge the battery while sailing?

Yes, routed through the BMS charge path. Regen from a propeller under sail runs 50 to 250 W on a small craft, and it must be gated off when cell temperature is below freezing or the pack is already full.

What certifications should I ask a battery supplier for?

IEC 62133-2 at cell level, IEC 62619 or UL 1973 at pack level with test reports rather than a self-declaration, UN 38.3 for transport, and documentation showing the installation complies with ABYC E-13 or ISO 16315. Commercial vessels additionally need class society approval.

How long will a marine lithium pack actually last?

For LFP, 4,000 to 6,000 cycles to 80% capacity at 0.5 C and 25 °C, but treat 10 to 12 years as the realistic service life because bilge temperatures accelerate calendar ageing. Track DC impedance as well as capacity, because impedance drift is the earlier warning.

A custom battery solution for an electric outboard is a systems problem before it is a cell problem. Get the bus voltage right, size from measured power rather than marketing range, keep current density sane, and seal the enclosure properly. The packs that fail early are almost never the ones specified conservatively.


Further Reading

References


Similar Posts