In this project, I helped a friend planning, sourcing and assembling a relatively big battery to power a boat with an electric motor.
This project came along when we concluded that the currently available solutions are far too expensive for leisure use.
We started by laying out our requirements for the pack:
– Water splash-resistant
– Lower explosion / violent fire risk
– High endurance (high cycle count)
– 48 Volt
– Delivering around 250A to be able to comfortably power a 10-12Kw motor
– 12-18 kWh capacity
– Stable voltage to avoid the power loss effect when operating the boat at lower battery SoC
– Low amount of modules to avoid having to put too many cells together and make the project more accessible
– Information on the pack and on a phone / tablet to be able to monitor it
– Physical fit in a linder aluminum boat (pic 1)
– Being ready by the end of winter to be able to use it in the summer

We reasearched the different types of battery chemestries and ended up with this table “LiFePO4 vs NMC vs NCA”:
Advantage | Drawbacks | |
LiFePO4 | – Long lifespan (2000-10000 cycles) – Very safe, low risk of thermal runaway – Stable voltage and high thermal stability – Environmentally friendly (non-toxic materials) | – Lower energy density than NMC – Heavier and bulkier for the same capacity – Low charge / discharge rate (1C optimally, 2C works) |
NMC | – High energy density (more capacity per size/weight) – Balanced lifespan and power output – Used in EVs and high-performance applications | – Less stable than LiFePO4, can overheat – Shorter cycle life (around 1000-2000 cycles) |
NCA | – Highest energy density among the three – High power output and efficiency – Commonly used in high-performance EVs (e.g., Tesla) – Lightweight compared to LiFePO4 and NMC | – Lower lifespan (typically 500-1000 cycles) – Less thermal stability, higher risk of overheating – Expensive due to high cobalt and nickel content |



Although NMC and NCA are very compelling options, we ended up going with LiFePO4 because of its stability and safety especially as we had no prior experience. The weight penalty didn’t seem like an issue as boats are generally heavy and slow at reacting anyways. We started looking for specific cells with the configuration: 16S1P giving us 51V when fully charged (a “48V” motor actually runs 52v-42v because of the big voltage differences through the charge range of conventional li-ion batteries). We then chose 320 ah cells to have a total capacity of 16 kWh.
We chose to build 2x 24V packs that we would wire in series to be able to fit the packs in the alu boat. We later on regretted not going with 2x 48v 160ah instead of 2x 24v 320ah as it would be easier to wire and offered more flexibility.
For the BMS, we went with 2x Daly BMS, each made for an 8s setup (24v). We chose Daly as they are pretty cheap and widely used which provides a big community of DIY’ers. We chose a model with a max power draw of 250A, power(W)=voltage(V)*current(A), 48V*250A= 12000W, which means we could power the 10kW motor with overhead. It also meant discharging our cells at under 1C, keeping them in ideal conditions.
We contacted chinese suppliers on Alibaba and, after finding the best price, placed an order for our cells, bus bars and BMS.


When we recieved the batteries, we planned the cabling between both batteries, bought simple aluminum boxes and started assembling the cells, measuring instruments and BMS . We used Anderson connectors as they are available in a size adapted to 250A of current, are easy to operate with gloves / wet hands and use crimp press for mounting on cables, thus not requiring soldering(important for us as it was not duable with our equipement to solder the big cables needed for a 250A system).



We used this first version of the battery pack throughout summer 2022 until the electric outboard broke (that’s a whole story in itself). The battery system worked as expected—we didn’t experience overheating issues, and none of the components failed. Even though we considered that a pretty big win, we knew there was room for improvement.
In the meantime, we had to accept that achieving planing with the electric boat wasn’t feasible, so we shifted our focus to a pontoon boat. This change in hull type gave us the opportunity to redesign the battery system from the ground up, allowing for better integration into the boat.
We decided to build the battery as one large box and use it as a table in the middle of the boat. During further research, we discovered that compressing the battery cells would increase their longevity. To achieve this, we used two wooden plates and four threaded rods to compress the cells. The rods also served as mounts for the BMS on each 8S pack.
We built the V2 battery with three main components: two 8S 24V 320Ah packs with BMS and an “electronics board.” The electronics board included the serial wiring of the two packs, ports for external connections, and two 220V-to-24V chargers. This setup allowed us to integrate all the electronics under the table on the pontoon boat while still enabling the battery packs to be removed without requiring tools to lift them out of the boat.
The box containing the cell packs and electronics board is a completely waterproof, thick plastic box, significantly improving waterproofing.
We also added various converters and inverters to the electronics board to provide 220V, 12V, and 5V outputs for auxiliary peripherals.





Looking back, choosing LiFePO4 cells was the right decision, as our limited experience led us to make mistakes that could have been more critical with other chemistries. For example, we initially failed to compress the cells and had inappropriate waterproofing for the environment in the first version.
If I could start from scratch, I would ensure that each battery pack was built to the desired voltage from the beginning, avoiding the need to wire packs in series during use. This would also provide greater modularity, allowing for easier addition or subtraction of capacity.
Leave a Reply