Home | Projects | Notes > Battery Management Systems (BMS) > State of Charge (SOC)

State of Charge (SOC)

What is SOC

State of Charge (SoC) represents how much energy is currently stored in a battery relative to its usable capacity, typically expressed as a percentage (0-100%). From a system and user perspective, SOC is analogous to a fuel gauge in a vehicle. It is the most intuitive indicator of remaining battery energy.

A common conceptual definition is:

However, the key nuance is that usable capacity is not a fixed value. It varies with:

• Battery aging and degradation (capacity fade)

• Temperature

• Charge/discharge rate (C-rate)

• Operating voltage limits defined by safety requirements

Because of this, SOC is not directly measurable and must always be estimated.

Why is SOC Important?

SOC is a core variable in almost every Battery Management Systems (BMS) decision. Accurate SOC estimation is critical for:

• User Experience

  • Providing reliable remaining runtime or driving range

  • Preventing unexpected shutdowns

  • Avoiding misleading SOC jumps or drift

• System Control & Safety

  • Enabling or disabling charge/discharge based on safe operation limits

  • Preventing overcharge and overdischarge

  • Coordinating with power electronics (inverters, chargers, DC/DC converters)

• Energy Management & Optimization

  • Load balancing and power limiting

  • Optimal charge scheduling

  • Integration with higher-level systems (EMS, vehicle controllers, grid controllers)

In short, a BMS with poor SOC estimation is fundamentally unreliable, regardless of how good the hardware is.

Why is SOC Estimation Hard?

Unlike voltage or current, SOC cannot be measured with a sensor. The difficulty comes from:

• Non-linear battery behavior

• Strong dependency on temperature and load

• Hysteresis effects

• Cell-to-cell variation

• Aging-related parameter drift

As a result, all SOC values are model-based or inference-based estimates, not ground truth.

Common SOC Estimation Methods

Voltage-Based Estimation (OCV Method)

Principle

SOC is estimated using the relationship between Open-Circuit Voltage (OCV) and SOC.

How it Works

• Measure battery voltage at rest (no load, sufficient relaxation time)

• Map voltage to SOC using a predefined OV-SOC curve.

Advantages

• Simple to implement

• No current sensor required

• Useful for initialization or recalibration

Limitations

• Requires long rest periods to be accurate

• Highly temperature-dependent

Typical Use Cases

• Initial SOC estimation at startup

• Periodic correction for other algorithms

Coulomb Counting (Current Integration)

Principle

SOC is calculated by integrating current over time. (Charge current: +)

Advantages

• High resolution and fast response

• Accurate over short time scales

• Works under dynamic load

Limitations

• Accumulates error over time (drift)

• Sensitive to current sensor offset and noise

• Requires a reliable initial SOC

Typical Use Cases

• Real-time SOC tracking

• Combined with correction mechanisms (OCV, model-based)