Power Station Display Accuracy: SOC Percentage and Voltage Readings
Volume I · May 2026 · 983 words
The state-of-charge percentage displayed on a portable power station is an estimate — not a measurement. Under some conditions, the estimate can be off by 20% or more, leading users to believe they have more (or less) runtime than they actually do. Understanding how the estimate is produced, when it is reliable, and when to trust voltage instead is essential for managing energy during an outage.
How SOC Estimation Works
The BMS estimates state of charge using one or both of two methods:
Coulomb Counting
The primary method. The BMS integrates current flow over time — counting coulombs (ampere-seconds) entering and leaving the battery. Starting from a known reference point (full charge), each coulomb drawn from the battery represents a known quantity of energy. The BMS subtracts this from the nominal capacity and displays the result as a percentage.
The problem: coulomb counting accumulates error. Every current measurement has a small offset (typically 0.1–1% of the measurement range). Over multiple partial charge-discharge cycles — the typical usage pattern for emergency power stations — the accumulated error can reach 10–20% before recalibration. The display might show 30% remaining when the battery is actually at 15%, or 80% when it is at 95%.
This is why manufacturers recommend a full discharge-recharge cycle every few months: it recalibrates the coulomb counter at both 0% and 100% endpoints. Between calibrations, the SOC display becomes progressively less accurate.
Voltage-Based Estimation
The backup method. The BMS measures the battery's open-circuit voltage and looks up the corresponding SOC from a pre-programmed discharge curve. This is accurate when the battery has been at rest (no load, no charge) for ≥ 30 minutes — enough time for voltage to settle after the electrochemical relaxation period. Under load, voltage sags; during charging, voltage rises. Using voltage to estimate SOC while current is flowing produces large errors.
Voltage-based estimation is most useful at the extremes: a LiFePO₄ battery at 3.20 V per cell is near empty (~10% SOC); at 3.40 V per cell, it is near full (~90% SOC). The middle of the discharge curve (3.25–3.35 V) is nearly flat — a 0.01 V difference corresponds to a 10–20% SOC difference, making precise estimation impossible in this range.
When the Display Is Reliable
| Condition | Display accuracy |
| Immediately after a full charge | Accurate (± 2%). The coulomb counter was just reset to 100%. This is the most reliable reading. |
| During a continuous discharge from 100% | Accurate (± 5%). Error accumulates slowly over hours, not minutes. A single continuous discharge does not accumulate significant drift. |
| After multiple partial cycles without recalibration | Unreliable (± 10–20%). Error has accumulated over days or weeks of partial use. This is the most common state for emergency-use power stations. |
| Under heavy load (≥ 80% inverter rating) | Unreliable. Voltage sag under load makes the displayed SOC drop temporarily. When the load is removed, SOC "recovers" by several percent. |
| At very low SOC (< 10%) | Unreliable. The discharge curve steepens, making estimation more sensitive to small voltage errors. The unit may shut down earlier than the display suggests. |
Practical Implications
During an outage, treat the SOC display as an optimistic estimate with a 15–20% margin. If the display shows 30% remaining, plan as if you have 10–15%. This margin accounts for coulomb counting drift, voltage sag under load, and the BMS cutoff reserve (the 5–10% of capacity the BMS reserves to prevent over-discharge damage — the unit shuts down at "0%" displayed even though the cells still hold charge).
A more reliable indicator than SOC percentage is runtime tracking: note the time when you began discharging and estimate runtime from the known load. If a 768 Wh unit powers a 100 W load, the theoretical runtime is 7.7 hours (100% to 0%). After 4 hours at a steady 100 W, you have consumed 400 Wh — approximately 52% of nominal capacity, regardless of what the display shows. Your own energy accounting is more reliable than the BMS estimate between calibrations.
Recalibration Procedure
Every 3–6 months, perform a full calibration cycle: discharge the unit to automatic shutdown (0% displayed), then recharge to 100% without interruption. The BMS recalibrates at both endpoints. After calibration, the SOC display returns to ± 2–5% accuracy. If the display shows unusual behavior — rapid fluctuations, jumps of 10% or more, or failure to reach 100% after extended charging — the BMS may have a fault; contact the manufacturer.