Battery Capacity Testing: Verifying Your Power Station's Actual Capacity

Volume I  ·  May 2026  ·  861 words

The capacity printed on the box — "768 Wh" — is the capacity when new, measured under specific conditions. A unit that is 3 years old, has been stored at high temperature, or has accumulated hundreds of cycles may have significantly less. Testing actual capacity tells you what you really have available during an outage.

Method 1: Resistive Load Test (Most Accurate)

Connect a known resistive load — an incandescent lamp, space heater on low, or a dedicated load tester — to the AC output. Record the starting time and the load wattage (verify with a Kill-a-Watt or the power station's own display). Run until the unit shuts down (0% displayed).

Actual Wh = Load (W) × Runtime (hours)

Example: a 100 W incandescent bulb runs for 2.8 hours before shutdown. Actual capacity = 100 × 2.8 = 280 Wh. If the unit is rated for 288 Wh nominal (233 usable AC), 280 Wh suggests the battery is healthy — the usable capacity is close to nominal after accounting for inverter losses.

This is a destructive test — it fully discharges the battery, consuming one cycle. Perform it once when the unit is new to establish a baseline, and repeat annually to track degradation. A 5–10% reduction from baseline over 3 years is normal; a 20%+ reduction suggests a problem (defective cell, BMS fault, or abuse).

Method 2: Energy Accounting (Non-Destructive)

During normal use, track energy consumption with a power meter or the unit's display. Start from a known SOC (fully charged after recalibration is ideal), use the unit normally, and note the SOC before recharging. The energy consumed divided by the SOC change gives an estimate of full capacity.

Estimated capacity = Energy consumed (Wh) ÷ (SOC_start − SOC_end) × 100%

Example: from 100% to 40% SOC, the unit delivered 180 Wh. Estimated capacity = 180 ÷ 0.60 = 300 Wh. Repeat over multiple partial cycles and average the results. A minimum of 3 cycles is required for a reliable estimate, because SOC display drift (see SOC accuracy) introduces error in each individual measurement.

Method 3: DC Load Test (For Advanced Users)

If the unit has a regulated 12 V DC output, connect a known DC load and measure runtime. This bypasses the inverter, measuring battery capacity more directly — inverter losses are excluded. A 12 V, 5 A load (60 W) running for 4.5 hours = 270 Wh DC capacity. Compare to the AC load test result; the difference is the inverter loss. If AC capacity is significantly lower than DC capacity, the inverter may be degrading.

Interpreting Results