Battery degradation rate is a function of temperature history over the cell's lifetime, not just peak temperature in any single event. This is an obvious statement when written out, but the design practice of many pack teams treats ambient temperature as a steady-state parameter rather than a time-varying input that shapes cumulative aging. In climates with large seasonal temperature swings, this matters significantly for cycle life prediction and protocol optimization.
For two-wheeler packs in India, the ambient temperature distribution across a year is not centered on 25 degrees Celsius. In northern cities like Delhi, Jaipur, and Lucknow, the range spans roughly 8 degrees Celsius in December to 44 degrees Celsius in May. In southern and central cities including Bengaluru, Chennai, and Hyderabad, the range is narrower but concentrated in the 22 to 40 degrees Celsius band with essentially no cold extremes. This means the pack sees a distinctly different cumulative heat dose depending on the season, and a protocol designed around a single representative ambient will be mismatched for a large fraction of the operating year.
What cumulative heat dose does to SEI growth
SEI growth rate on the graphite anode follows Arrhenius kinetics. The rate constant for SEI growth roughly doubles for every 8 to 12 degree Celsius increase in temperature, depending on the electrolyte formulation. This is not a slight sensitivity. A cell cycling at 40 degrees Celsius is aging its SEI approximately two to three times faster than the same cell cycling at 25 degrees Celsius.
For a two-wheeler pack in Delhi, consider the seasonal breakdown: approximately three months of heavy summer cycling in the 35 to 44 degree range, three months of moderate shoulder season in the 25 to 35 degree range, and six months of lower-temperature cycling from December through March. The total SEI growth in the three summer months is comparable to or greater than the remaining nine months combined, because of the exponential temperature sensitivity. The pack ages fastest in summer and almost rests its SEI during winter.
This has protocol implications. In summer, the charge protocol should minimize unnecessary thermal stress: lower charge cutoff voltage if SoC utilization allows it, reduce peak charge rate at high SoC, charge during the coolest part of the day if usage patterns allow it. In winter, the protocol has more thermal headroom and can afford to charge faster without accumulating disproportionate SEI growth. A static single-protocol approach imposes the same rate year-round, which is either too aggressive in summer or too conservative in winter.
High humidity and electrolyte effects
Coastal and monsoon-season conditions add a complication that is not present in northern-hemisphere battery data: prolonged high humidity exposure combined with high temperature. For most production cells, humidity inside the cell itself is controlled during manufacturing (lithium-ion cells are assembled in dry rooms). The concern is not water ingress through the cell housing in a properly sealed cell, but rather the thermal conductivity and heat rejection capability of the pack housing under high-humidity conditions.
The surface thermal coefficient for natural convection depends on the temperature differential between the cell surface and the surrounding air. When ambient temperature is already high and humidity is high, the air's ability to absorb heat from the pack surface is reduced compared to dry conditions at the same temperature. This effect is secondary compared to the direct temperature effect on degradation, but it compounds the challenge of managing pack temperature in a coastal Indian summer without active cooling.
For pack designers, the practical implication is that passive cooling systems should be sized for the worst-case combined condition (high ambient temperature plus high humidity) rather than for a temperature-only baseline. The thermal path from cell to ambient air is less effective under those conditions, and the margin that exists at 35 degrees Celsius in dry conditions may not be present at 35 degrees Celsius with 85% relative humidity.
The protocol parameterization challenge
Designing a charge protocol that is appropriate for the full Indian ambient temperature distribution requires either a continuously adaptive protocol that adjusts in real time, or a set of seasonal protocol variants that are selected based on ambient temperature range.
The continuously adaptive approach has the theoretical advantage of optimally matching protocol to conditions at all times. It requires the BMS to read ambient temperature, estimate cell temperature from a model, and adjust the current profile accordingly. This is achievable with modern BMS microcontrollers if the thermal model is simplified enough for real-time execution. The practical challenge is that the control logic becomes more complex and requires more thorough validation across the full operating range.
The seasonal variant approach is simpler to implement and validate. Define three protocol tiers: cool season (ambient below 20 degrees Celsius), moderate season (20 to 32 degrees Celsius), and hot season (above 32 degrees Celsius). Parameterize each tier from a thermal model of the cell, allowing higher peak charge rates in the cool season and stepping down in the hot season. The BMS selects the tier based on ambient temperature at charge session start. This captures most of the degradation variance associated with temperature seasonality with relatively simple implementation.
Validating against Indian ambient data
One issue we have encountered when advising pack engineering teams is that cell characterization data often comes from laboratory testing conducted at 25 degrees Celsius with occasional tests at 0 and 45 degrees Celsius as boundary conditions. The intermediate range, particularly the 30 to 40 degree band where much of the Indian operational year sits, is often sparsely characterized. Models built from 25 and 45 degree data points interpolate poorly through the 30 to 40 degree range because the degradation mechanisms do not have linear temperature dependencies.
We would push back on the "test at extremes and interpolate" approach for this application. Test at 30 and 37 degrees Celsius as representative of the Indian nominal condition. The data points that cover the actual operating distribution are more informative for protocol design than the theoretical extremes. This is a simple adjustment to cell characterization protocol that significantly improves the relevance of the resulting model for Indian market applications.
The long-term consequence of ignoring seasonal variance
Over a three-year two-wheeler pack lifetime, the cumulative difference in SEI growth between a thermally-adaptive protocol and a static worst-case protocol is measurable in capacity retention terms. In our internal modeling of a 2.2 Ah 18650 cell cycled with a representative Indian temperature distribution, a static 1C protocol throughout the year results in 14 to 18% more cumulative SEI growth in the first 500 cycles than a seasonally-adapted protocol that uses 1.5C in the cool months and 0.8C in peak summer. Both protocols stay within safe thermal bounds; the difference is how much of the available thermal margin they utilize when conditions allow and how carefully they protect the cell when conditions are adverse.
That 14 to 18% difference in cumulative SEI growth translates to meaningfully different capacity retention at the end of the pack warranty period. It is also the kind of difference that is invisible to a user until the pack is noticeably degraded, at which point the protocol that caused it is already two years in the past. Designing the protocol with the full ambient distribution in mind from the start is how you avoid building that margin in retroactively as a warranty issue.