Battery state of charge vs voltage: why they differ

Voltage describes a condition; percentage is an estimate

Voltage is a measurement at a particular point and moment. State of charge (SOC) estimates how much charge remains relative to the battery’s usable or configured capacity. They answer different questions. Turning one voltage reading into a precise percentage discards information about the battery chemistry, load, temperature and recent charging.

Texas Instruments describes the flat voltage-versus-SOC characteristic of LFP cells in its battery-monitoring white paper. This is why a generic “48 V battery percentage chart” is a poor substitute for a battery-specific measurement system. A nominal system voltage is not a full-charge calibration.

Find the device behind the percentage

A BMS may report its own estimate. A shunt-based monitor measures current flowing into and out of the battery and tracks that movement over time. An inverter may provide another SOC value whose method depends on its model and configuration. Three percentages on three screens need not have the same source.

Victron’s SmartShunt operation manual explains current integration and periodic synchronization. Its troubleshooting guidance identifies capacity settings, current measurement and synchronization problems as reasons for a misleading percentage. Those principles explain the question to ask; they are not universal settings to copy into a different battery.

Before trying to correct a discrepancy, label each display: “BMS pack 1,” “inverter,” or “separate battery monitor.” Then compare readings captured at the same time. An old value and a fresh value are not a useful calibration test.

A capacity mistake changes the whole story

For a multi-pack bank, also check scope. A percentage from one communicating pack cannot establish the condition of a second silent pack. In Sunstead, directly reported pack values and inverter-terminal readings are distinguished; cell detail requires a supported BMS connection that actually supplies it.

Treat runtime as a conditional calculation

Suppose a battery has 4 kWh available above a chosen reserve and the measured draw is 0.8 kW. Dividing gives five hours if that draw stays constant. A pump starting, a refrigerator cycling, new sunlight or a change in usable capacity changes the answer. It is a planning estimate, not a promise of five hours of service.

Sunstead leaves unsupported measurements unavailable. It does not invent cell voltages from the inverter voltage or infer a full bank’s energy capacity from a single instantaneous reading. A missing runtime estimate is more useful than a confident number based on an unknown battery.

When the numbers disagree, collect these facts

• The source of each SOC value and its timestamp.
• The exact battery model, chemistry and configured capacity.
• Whether the battery was charging, discharging or resting.
• Whether every installed pack is represented by a fresh reading.
• Whether the monitor recently restarted or completed its manufacturer-defined synchronization.
• Any simultaneous BMS warnings, temperature readings or reported cell differences.

Bring that record to the battery manufacturer’s troubleshooting instructions. Do not force a display to 100% just to make two screens agree, or change charging limits from a generic percentage chart. The useful goal is to identify the source of the disagreement.

For the rest of the screen, see how to read solar, load and battery flow together. If the BMS is missing entirely, use the monitoring compatibility checklist.