State of Charge (SoC)

State of Charge (SoC) is the equivalent of a fuel gauge for a battery, representing the current electrical energy stored as a percentage of its maximum usable capacity (0% = empty, 100% = full). Unlike a fuel tank, the 'full' capacity degrades over time, making precise measurement a complex estimation problem rather than a simple volume reading.

Direct measurement is impossible; SoC must be inferred algorithmically. Common methods include:

• Coulomb Counting: Integrating current flow over time, prone to drift error accumulation.
• Open Circuit Voltage (OCV): Measuring voltage at rest; requires long relaxation periods.
• Kalman Filtering: A predictive state estimator that fuses voltage, current, and temperature sensor data with a battery model to correct for sensor noise and model inaccuracies.
• Extended Kalman Filters (EKF) are widely used to handle the non-linear voltage-SoC relationship of lithium-ion cells.

The usable SoC window is deliberately restricted by the Battery Management System (BMS) to preserve longevity. A displayed 0% often corresponds to a true chemical SoC of ~5-10%, and 100% to ~90-95%. Operating a lithium-ion cell at extreme voltage limits (true 0% or 100%) accelerates Solid Electrolyte Interphase (SEI) growth and irreversible capacity fade.

SoC is the primary input variable for Model Predictive Control (MPC) in EV charging optimization. Algorithms require precise SoC to:

• Calculate the energy required to reach a target SoC by a departure time.
• Determine the available flexibility for Demand Charge Management or Vehicle-to-Grid (V2G) dispatch.
• Prevent Depth of Discharge (DoD) violations that would accelerate battery degradation.

SoC is a short-term, reversible metric (energy available right now). State of Health (SoH) is a long-term, irreversible metric representing the battery's degradation relative to its Beginning of Life (BoL) capacity. A battery at 100% SoC may only hold 80% of its original energy if the SoH has degraded to 80%. SoH directly constrains the maximum achievable SoC.

Automotive applications typically require SoC estimation error < 5% to ensure reliable range prediction. Key error sources include:

• Current sensor offset: A tiny DC bias in the current sensor integrates into a massive Coulomb counting error over hours.
• Temperature effects: Internal resistance and available capacity change dramatically with temperature, skewing voltage-based estimates.
• Cell imbalance: In a series pack, the weakest cell determines the pack SoC; balancing circuits must be accounted for in the estimation algorithm.