Introduction: The Hidden Health of Your Battery Pack
In the previous instalment of Autoven's BMS series, we explored methods for accurate State of Charge (SoC) estimation — including the Extended Kalman Filter (EKF), which has emerged as the industry-standard algorithm for real-time SoC tracking. Now we go one layer deeper.
Leading BMS manufacturers in India are developing EV battery management system solutions with advanced SoH estimation, so that battery degradation can be tracked and managed over the life of the pack.
SoC tells you how full the battery is right now. State of Health (SoH) tells you how much the battery can hold compared to when it was new. It is the long-term health metric that determines actual range, warranty validity, resale value, and ultimately when the pack needs replacement. For Indian EV OEMs, fleet operators, and BaaS providers managing hundreds of vehicles across varying climates and duty cycles, SoH is not a dashboard number — it is a business-critical signal.
This article is the second instalment in Autoven's BMS accuracy series. Part 1: Methods for Accurate SoC Estimation in Electric Vehicle Batteries. Part 3: Enhancing SoP and DTE Estimations for Li-ion Battery Packs.
What Is State of Health (SoH)?
State of Health (SoH) is a percentage that represents the current maximum charge capacity of a battery pack relative to its original rated capacity. A new battery starts at 100% SoH. As it ages — through cycling, heat exposure, and the electrochemical processes inherent to lithium-ion chemistry — SoH declines.
Example:
Original rated capacity : 60 kWh
Current maximum capacity: 48 kWh (after 3 years of use)
SoH = (48 / 60) × 100 = 80%
At 80% SoH, this EV has 20% less range than when new — even at 100% charge.
SoH is not a single physical property that can be measured directly — it must be estimated by the BMS from observable signals such as voltage, current, temperature, and charge history. This estimation challenge is central to advanced BMS design.
The practical consequences of SoH decline are significant and cumulative. An EV that once offered 120 km of range on a full charge will deliver only 96 km at 80% SoH — and this decline happens gradually, often without drivers noticing until the gap is substantial. For Indian fleet operators running EV auto-rickshaws on 8–10 hour daily duty cycles, an untracked SoH decline can silently erode per-vehicle economics month by month.
Why Accurate SoH Estimation Matters
Accurate SoH estimation is not just a technical nicety — it has direct financial and safety consequences for every stakeholder in the EV value chain.
For EV OEMs
- Premature replacement prevention — Without accurate SoH data, warranty teams replace batteries based on customer complaints rather than actual condition, leading to unnecessary costs
- Design feedback — Real-world SoH decline patterns reveal which cell chemistries, pack configurations, and charging profiles degrade fastest, informing future product improvements
- Warranty policy calibration — Data-backed SoH thresholds allow OEMs to set warranty terms confidently — not conservatively padding them at customer expense
For Fleet Operators
- Predictable range planning — Dispatchers need to know the actual range of each vehicle in the fleet, not the nameplate range. SoH-corrected range estimates enable reliable scheduling
- Planned maintenance over emergency failures — A BMS that tracks SoH trends can predict replacement 3–6 months in advance, enabling planned procurement instead of emergency replacement
- Asset financing advantage — Verified SoH data from AV BMS strengthens EV loan applications with banks and NBFCs, as lenders can assess residual asset value objectively
For BaaS Providers
- Per-battery economics — BaaS business models depend on knowing the remaining useful life of each battery asset. Inaccurate SoH makes asset lifecycle planning impossible
- Swap station optimisation — SoH data determines which batteries should be retired from active swap inventory vs. repurposed for stationary storage applications
Inaccurate SoH creates three specific failure modes: premature replacement (replacing batteries that still have usable life, wasting thousands of rupees per pack), unexpected failures (running batteries past safe SoH thresholds until they fail in the field, causing downtime and safety risks), and inefficient warranty planning (OEMs over-provisioning warranty reserves because they cannot predict actual failure rates). All three are directly preventable with an accurate SoH estimation system.
Methods for Estimating State of Health
The EV battery management system uses a combination of direct measurement, mathematical modelling, and historical data analysis to estimate SoH. No single method is sufficient in all conditions — the most accurate BMS implementations combine multiple approaches into a unified estimation pipeline.
1. Capacity-Based Estimation
The most straightforward approach: measure the total energy a battery delivers during a complete discharge and compare it to the original rated capacity. Capacity-based estimation is highly accurate when a full charge-discharge cycle is available, making it ideal for periodic calibration during overnight fleet charging or scheduled maintenance.
Its limitation is real-time applicability — most vehicles never complete a full calibration cycle in normal operation. Partial discharge data can be extrapolated, but with reduced accuracy. Capacity-based estimation works best as a periodic ground-truth calibration that corrects model-based estimates over time.
2. Internal Resistance Measurement
As a Li-ion battery ages, its internal resistance increases due to electrolyte decomposition, SEI (Solid Electrolyte Interphase) layer growth, and electrode degradation. Internal resistance is a reliable long-term SoH proxy — a pack at 80% SoH typically shows 20–40% higher resistance than a new pack of the same chemistry.
The BMS measures resistance by observing the voltage drop under a known current pulse — a technique practical enough to run continuously in the background of normal vehicle operation. In India's high-temperature conditions, thermal compensation is critical because resistance also varies with temperature, and a hot battery may appear healthier than it actually is without proper correction.
3. Model-Based Estimation (EKF and Machine Learning)
The most powerful approach for real-time SoH estimation. Model-based methods use a mathematical representation of the battery's electrochemical behaviour, continuously reconciled against live sensor data by an adaptive algorithm.
- Extended Kalman Filter (EKF) — Dynamically estimates battery states by combining a physics model with real-time sensor corrections. Handles measurement noise, model uncertainty, and changing operating conditions simultaneously
- Unscented Kalman Filter (UKF) — More accurate than EKF for highly non-linear battery behaviour, at greater computational cost
- Machine Learning estimators — Neural networks trained on large datasets of charge/discharge cycles can identify SoH patterns that purely model-based approaches miss, especially in batteries with unusual ageing histories
- Electrochemical Impedance Spectroscopy (EIS) — Emerging technique that characterises the battery's internal impedance spectrum to infer detailed degradation mechanisms — currently more common in lab settings but moving towards vehicle-grade implementations
| Method | Real-Time? | Accuracy | Best For |
|---|---|---|---|
| Capacity-Based | No (periodic) | High (full cycle) | Periodic calibration / fleet overnight charging |
| Internal Resistance | Yes | Medium | Continuous background monitoring, trend tracking |
| EKF Model-Based | Yes | High | Production BMS — standard real-time estimation |
| Machine Learning | Yes (inference) | Very High | Advanced BMS with sufficient training data |
Key Strategies to Improve State of Health
Improving SoH is not only about measuring it accurately — it is about proactively managing the factors that drive degradation. The four strategies below are the primary levers available to BMS designers and fleet operators to extend battery lifespan in Indian operating conditions.
Temperature management · Optimal charging and discharging currents · Appropriate Depth of Discharge (DoD) · Regular cell balancing. These four factors together determine the rate of SoH decline. A BMS that actively manages all four can extend usable battery lifespan by 30–50% compared to an unmanaged pack.
Strategy 1: Temperature Management
Temperature is the single most powerful driver of SoH degradation. Li-ion batteries age fastest at both extremes — cold temperatures accelerate lithium plating during charging, while high temperatures accelerate electrolyte decomposition and SEI layer growth. Maintaining operating temperature between 20°C and 30°C significantly extends battery lifespan.
For Indian EVs, high heat is the dominant concern. In Chennai, Rajasthan, or Maharashtra, ambient temperatures regularly exceed 40°C from March to June. A battery pack sitting in direct sunlight in these conditions can reach internal temperatures of 50–55°C, conditions under which SoH degradation accelerates dramatically.
- Battery Thermal Management System (BTMS) — Active cooling (liquid or forced-air) maintains pack temperature within the optimal range during operation and charging
- Thermal pre-conditioning — Cooling the pack before a fast-charging session prevents the combined thermal stress of charge heat plus ambient heat
- Parking and storage practices — For fleet operators, covered or shaded parking reduces calendar ageing during off-shift hours
- Temperature-aware charging limits — The BMS reduces charging C-rate dynamically when pack temperature exceeds thresholds, trading charge time for longevity
Tesla uses liquid cooling loops in its EV BMS to maintain even temperature distribution across cells. For Indian cost-sensitive 2-wheeler and 3-wheeler segments, leading BMS manufacturers are designing optimised air-cooled systems that balance thermal performance with cost, weight, and reliability in dusty Indian road conditions.
Strategy 2: Optimal Charging and Discharging Currents
The rate at which a battery charges or discharges — expressed as a C-rate (multiples of the battery's capacity) — directly impacts SoH. High C-rate charging is the fastest path to SoH degradation. Fast charging at 2C or above generates significant heat, promotes lithium plating on the anode, and induces mechanical stress on electrode structures through rapid lithium intercalation and de-intercalation.
For regular daily charging, 0.5C to 1C is the ideal charge rate. At 0.5C, a 60 kWh pack charges from empty to full in approximately 2 hours — fast enough for overnight fleet charging, slow enough to minimise electrode stress. Fast charging should be reserved for urgent situations, not routine operations.
Autoven's AV BMS dynamically limits charging and discharging C-rates based on real-time temperature, SoC, and SoH. When the pack is hot, cold, or aged, the BMS automatically reduces the current ceiling — protecting the battery without requiring any manual intervention from the driver or fleet operator.
Strategy 3: Appropriate Depth of Discharge (DoD)
Depth of Discharge is the percentage of the battery's capacity used in a single discharge cycle before recharging. The deeper the discharge, the greater the electrode stress, and the faster the SoH declines. Limiting DoD is one of the most effective — and purely software-implementable — strategies for extending battery lifespan.
The industry recommendation is to limit DoD to 10%–90% of capacity for daily operation, avoiding both deep discharge (below 10%) and prolonged storage at full charge (above 90–95%). For high-value commercial EV applications where longevity is paramount over daily range, some BMS configurations restrict DoD to 20%–80%.
Reduced DoD to : 60%
Estimated cycle life : 1,300 – 1,400 cycles
Range reduction : ~25% per cycle (60 kWh usable → 45 kWh)
Net lifespan gain : +300–400 additional cycles (30–40% more)
Instantaneous Cycle Count = (DoD / 80%) × (1/2)
Use this formula in BMS design to predict cumulative ageing from partial cycles.
For Indian fleet operators, this trade-off has a clear economic interpretation: a fleet vehicle that operates within 20%–80% SoC (60% DoD) will need battery replacement 30–40% later than one operated at full 80% DoD. At a battery replacement cost of ₹40,000–₹80,000 per pack for a 3-wheeler, the DoD strategy alone can save significant capex over a 5-year fleet lifecycle.
Strategy 4: Regular Cell Balancing
A lithium-ion battery pack is made of tens to hundreds of individual cells connected in series and parallel. Due to manufacturing tolerances, temperature gradients across the pack, and natural variation in ageing rates, cells in the same pack will develop slightly different voltages over time. Without correction, these imbalances compound — accelerating the weakest cells' degradation and constraining the entire pack's performance to the weakest-link standard.
Cell balancing prevents this by equalising voltage across all cells, ensuring they charge and discharge uniformly and age at a consistent rate. Two approaches are used in production BMS systems:
- Passive Balancing — Bleeds excess energy from stronger cells as heat via resistors, equalising voltage levels downward. Simple, low-cost, and widely used in cost-sensitive Indian EV segments. The trade-off is wasted energy and heat generation during balancing
- Active Balancing — Redistributes energy from stronger cells to weaker cells using inductors or capacitors, equalising voltage upward. More efficient, no energy waste, and significantly better for SoH — but more complex and expensive. Best suited for premium EVs, BaaS packs, and commercial vehicles where battery longevity directly impacts per-unit economics
Autoven's AV BMS supports both passive and active cell balancing configurations, selectable based on the vehicle application and cost requirements. Cell-level voltage monitoring at high sampling rates detects imbalance early — before it compounds into measurable SoH divergence — enabling corrective balancing before degradation accelerates.
Cycle Life and SoH: Planning for the Full Battery Lifecycle
SoH decline follows a predictable trajectory influenced by all four strategies above. Understanding this trajectory allows OEMs and fleet operators to model battery lifecycle costs, plan replacement schedules, and make data-driven procurement decisions — rather than reacting to unexpected failures.
The rate of SoH decline is not linear. Batteries typically show rapid initial capacity loss in the first 50–100 cycles (the "break-in" period), then slow to a more gradual decline through the bulk of their life, before accelerating again as they approach end-of-life thresholds. A well-designed BMS tracks this curve and can project the remaining useful life (RUL) of each pack.
For fleet operators, this projection is the foundation of maintenance planning:
- Packs projected to reach 80% SoH within 60 days are flagged for planned replacement before field failure occurs
- Packs still above 85% SoH after expected cycle count are identified as outliers — potentially revealing optimal charging profiles worth replicating fleet-wide
- Packs showing accelerated SoH decline are investigated for thermal, mechanical, or charging anomalies before they become safety incidents
Impact of SoH on EV Ownership in India
For Indian EV buyers — whether individual consumers purchasing a 2-wheeler or fleet operators deploying hundreds of 3-wheelers — SoH has direct implications for ownership confidence, resale value, and total cost of ownership.
Indian EVs face compounding SoH stressors that European or Chinese-calibrated BMS systems are not optimised for: 45°C+ ambient summers accelerating thermal degradation, overloaded vehicles (3-wheelers frequently exceed rated payload) driving higher discharge currents, inconsistent charging infrastructure with variable voltage and temperature, and multi-operator fleet vehicles accumulating high annual cycle counts. A BMS tuned for these realities produces meaningfully better SoH outcomes than an off-the-shelf import.
Accurate SoH data enables three critical ownership benefits:
- Maintenance planning before failures occur — SoH trend data gives weeks or months of warning before a battery needs replacement, enabling planned action rather than emergency response
- Resale value preservation — A verified SoH report from AV BMS gives used EV buyers confidence in the battery condition, supporting higher resale prices and faster transactions
- Warranty claim validation — OEMs can distinguish genuine manufacturing defects (premature SoH decline) from misuse (deep discharge abuse, high-temperature charging) using BMS logs, making warranty processes fair and data-driven
AV BMS by Autoven: Physics-Based SoH for Indian EVs
Autoven's AV BMS is designed around a first-principles, physics-based approach to SoH estimation — not lookup tables derived from lab testing at room temperature. This matters because real-world SoH behaviour in Indian conditions — extreme heat, heavy loads, inconsistent charging — diverges significantly from lab assumptions. A physics-based model that understands the electrochemistry of degradation produces accurate estimates across all operating conditions, including the ones no lab ever tested for.
AV BMS SoH Capabilities
- Continuous real-time SoH estimation using EKF algorithm updated at each measurement cycle
- Cell-level voltage and temperature monitoring — weakest cell always governs pack-level SoH
- Adaptive charging current limits that reduce C-rate dynamically with temperature and SoH
- Configurable DoD limits enforced at the BMS level — not dependent on driver behaviour
- Active and passive cell balancing support, configurable by application
- SoH trend transmission via AV IoT to Autoven Platform for fleet-wide lifecycle analytics
- Remaining Useful Life (RUL) projections that flag packs approaching replacement thresholds
- India-specific thermal derating profiles calibrated for 0–55°C ambient range
Protect Your Battery Investment with AV BMS
Autoven's AV BMS gives you accurate SoH tracking, proactive degradation management, and fleet-wide battery health visibility — purpose-built for Indian EV operating conditions. Talk to our engineering team to see how it applies to your platform.
Talk to Our Engineers →Conclusion: SoH Is the Long-Term Pulse of Your Battery
State of Health is the metric that connects battery chemistry to business outcomes. It determines actual range, drives warranty costs, sets resale value, and signals when replacement is needed. Every EV programme that ignores SoH is managing its most expensive asset on guesswork.
The good news: SoH decline is not inevitable at any particular rate. With the right BMS architecture and operational practices, Indian EV OEMs and fleet operators can significantly slow degradation, extend pack life, and extract more value from every battery investment.
- Implement accurate SoH estimation — Use model-based methods (EKF or equivalent) combined with periodic capacity calibration for the highest accuracy across the full battery lifecycle
- Manage temperature actively — A battery thermal management system tuned for Indian summer conditions is the single highest-ROI investment in battery longevity
- Limit charging C-rate for daily operations — Reserve fast charging for exceptions, not the daily routine; 0.5C–1C charging is the longevity default
- Enforce DoD limits at the BMS level — Do not rely on driver behaviour; the BMS should protect the DoD window automatically
- Balance cells consistently — A well-balanced pack ages uniformly; cell imbalance is one of the fastest paths to accelerated, irreversible SoH decline
In the final instalment, we explore State of Power (SoP) and Distance to Empty (DTE) estimations — how these two parameters translate battery health into real-time actionable guidance for the vehicle and driver.