At PowerFabb, our Umargam, Gujarat facility specializes in custom lithium battery packs for B2B clients across electric mobility and energy storage applications. As part of ongoing quality validation for a recent client project, we conducted side-by-side real-world testing of two ebike battery packs with identical 15Ah nominal capacity but different chemistries and configurations:
- NMC pack: 13S3P configuration using 5Ah cells (nominal voltage ~48.1V, ~722 Wh energy)
- LiFePO4 pack: 16S1P configuration using 15Ah cells (nominal voltage 51.2V, 768 Wh energy)
Both packs underwent standard bench capacity testing (low-current constant discharge), which confirmed delivery of near-full rated capacity (~15Ah) with no evident cell damage or significant imbalance.
Testing was performed under controlled conditions on the same ebike:
- Mostly flat routes in typical riding environment
- Constant speed of ~25 km/h
- Same speed controller configured to 38V limit
- Gentle throttle input to minimize aggressive acceleration
Observed Results:
- NMC pack: 40 km range, corresponding to ~13 Wh/km effective consumption
- LiFePO4 pack: 25 km range, corresponding to ~22 Wh/km effective consumption
This resulted in an approximately 60% shorter range for the LiFePO4 pack despite its slightly higher nominal energy content and theoretical advantages (flatter voltage curve, higher safe depth of discharge potential).
Key Observations:
- The discrepancy appeared under real pulsed-load conditions typical of ebike use (brief current spikes from starts, minor road variations, etc.), even at low average power.
- LiFePO4 ‘s characteristic flat discharge curve (maintaining ~3.2–3.3V per cell / ~51–53V pack for the majority of capacity) contrasts with NMC’s more gradual voltage decline.
- Cutoff was triggered by the BMS (not the controller), occurring prematurely on the LiFePO4 pack while bench tests at low current extracted full capacity.
- Pack weight difference (LiFePO4 heavier due to lower energy density) likely contributed a portion of the increased apparent consumption via higher rolling resistance.
- No anomalies were detected in static bench testing, underscoring the difference between constant low-load lab conditions and variable real-world ebike loads.
These findings illustrate how nominal specifications and slow-discharge bench results may not fully predict performance in pulsed-load applications like ebikes. The flat voltage profile of LiFePO4 , while beneficial in many scenarios, can interact with BMS low-voltage protection thresholds in ways that manifest differently under dynamic loads compared to NMC.
Next Steps at PowerFabb: To better understand the contributing factors—such as exact voltage sag behavior under load, per-cell dynamics during cutoff, potential minor imbalances, or BMS threshold interactions—we are planning additional controlled testing. This will include:
- Real-time per-cell voltage monitoring during rides
- Varied load profiles to capture more data points
- Repeated trials under identical conditions
We will share further observations in upcoming posts as more data becomes available.
PowerFabb remains committed to rigorous testing and transparent reporting in our custom lithium battery development process. If your project involves ebike, EV, or energy storage packs and you’d like to discuss custom manufacturing requirements, contact us.