Using terminal voltage to estimate battery capacity depends entirely on the cell chemistry, current load, and resting state of the system. While older lead-acid and certain lithium chemistries show a steady voltage decline during discharge, Lithium Iron Phosphate (LFP)—the most common chemistry in modern residential solar—features an exceptionally flat discharge curve. Between 20% and 90% capacity, an LFP cell’s voltage barely shifts by a few hundredths of a volt. To estimate usable capacity accurately using voltage alone, you must evaluate resting Open-Circuit Voltage (OCV) or track the voltage “knees” at the extreme ends of the cycle.
Quick Answer
Voltage readings can reliably estimate capacity only at the extreme ends of the charge cycle or after the battery has rested unpowered for 30 to 60 minutes. Because modern LFP solar batteries maintain an extremely flat voltage plateau across most of their capacity, resting Open-Circuit Voltage (OCV) or coulomb-counting shunts must supplement raw terminal voltage readings.
Terminal Voltage vs. Open-Circuit Voltage (OCV)
A common mistake in the field is measuring a solar battery’s DC terminals with a multimeter while the inverter is actively running and attempting to correlate that number to State of Charge (SoC).
Load Sag Effect:
Resting Open-Circuit Voltage (OCV) ──[ Load Applied ]──> Terminal Voltage Sags
V_terminal = V_ocv - (I * R_internal)
Charging Swell Effect:
Resting Open-Circuit Voltage (OCV) ──[ Solar Applied ]─> Terminal Voltage Swells
V_terminal = V_ocv + (I * R_internal)
Terminal voltage is dynamic:
- Under load (discharge): Current flowing through internal cell resistance, busbars, and terminal lugs pulls the terminal voltage downward (voltage sag).
- Under charge: Incoming solar current pushes terminal voltage higher than the cells’ true chemical potential.
To read true resting Open-Circuit Voltage (OCV), the battery must be electrically isolated (inverter off or DC breaker open) and allowed to rest. In lithium cells, internal electrochemical polarization takes between 30 and 60 minutes to relax fully, allowing surface charges to dissipate and the voltage to stabilize.
Chemistry Differences: NMC vs. LFP
How much capacity information voltage can reveal depends directly on cathode chemistry:
| Parameter | Lithium Nickel Manganese Cobalt (NMC) | Lithium Iron Phosphate (LFP) |
|---|---|---|
| Nominal Cell Voltage | 3.6V to 3.7V | 3.2V to 3.25V |
| Full Charge Target | 4.20V per cell | 3.65V per cell |
| Fully Depleted Target | 2.80V to 3.00V per cell | 2.50V per cell |
| Discharge Voltage Curve | Sloped (Continuous downward gradient) | Flat (Stable plateau across 70% of capacity) |
| Mid-Range Voltage Utility | Moderate to High (Voltage tracks capacity well) | Very Low (3.28V can be 30% or 70% SoC) |
In an NMC pack (such as a Tesla Powerwall 2), cell voltage steps down in a measurable gradient throughout the discharge cycle. A technician can measure pack voltage and estimate capacity within a reasonable margin.
In an LFP pack (such as Enphase IQ, EG4, or SOK), cells remain pinned between roughly 3.20V and 3.28V for almost the entire discharge. Because normal production variations, slight temperature shifts, and minor calibration offsets exceed the voltage difference between 40% and 75% SoC, estimating mid-range capacity purely from a voltage reading is impossible.
Using the “Knees” of the Voltage Curve
While mid-range voltage is uninformative in LFP systems, the voltage knees—the rapid exponential changes in voltage at the top 10% and bottom 10% of the charge cycle—provide precise capacity markers.
LFP Voltage Curve Profile:
Voltage
3.65V │ /─── Upper Knee (100% Full)
3.40V │ /
3.25V │──────────────────────────────/ Flat Plateau (Indeterminate SoC)
3.10V │ /
2.50V │___/ Lower Knee (0% Empty)
└──────────────────────────────────
0% 10% 90% 100% Capacity
- The Upper Knee (>3.40V per cell / >54.4V on a 48V pack): When charging, voltage climbs slowly through the plateau, then rises rapidly past 3.40V per cell. The instant voltage spikes sharply upward under a moderate charging current, the cell has reached its absorption ceiling and holds roughly 95% to 100% capacity.
- The Lower Knee (<3.10V per cell / <49.6V on a 48V pack): During discharge, the voltage holds steady until available lithium ions are nearly exhausted. Once a cell drops below 3.10V, it falls steeply toward the 2.50V cutoff. When this rapid downward curve begins, the pack is within its final 5% to 10% of usable energy.
To understand how battery management systems detect these transition zones, see Understanding the “Knee” of the Voltage Curve.
Step-by-Step Diagnostic Voltage Test
If your monitoring software is offline, inaccurate, or showing an error, use this field procedure to assess your pack’s state using voltage:
1. Isolate the Pack
Open the AC breaker feeding the hybrid inverter and switch off the solar DC isolator. If safe and supported by your hardware, open the DC circuit breaker or disconnect switch between the battery and the inverter.
2. Let the Chemistry Relax
Wait a minimum of 30 minutes. If you take a multimeter reading immediately after a high-current discharge or charge event, surface charge will skew the reading by several tenths of a volt.
3. Measure Total and Per-Cell Voltage
Set a calibrated digital multimeter to DC volts:
- Measure across the main positive and negative battery terminals.
- If individual cell monitoring pins or busbars are accessible, measure each cell in series.
4. Cross-Reference the Reading (16-Cell / 48V LFP Nominal Standard)
- >54.0V (>3.37V/cell): Fully charged (roughly 95%–100% resting capacity).
- 52.0V to 53.2V (3.25V to 3.32V/cell): Indeterminate operational range. The battery has usable capacity, but could sit anywhere from 25% to 85% SoC.
- <50.0V (<3.12V/cell): Approaching empty (roughly 5%–10% remaining capacity).
- <48.0V (<3.00V/cell): Depleted. The battery has reached its lower cutoff boundary.
If your meter reads healthy resting voltage across the terminals but your inverter displays an empty state or refuses to power up, review Why Your Battery Reports 0% but Still Has Voltage.
When Voltage Reads Artificially Low
If your resting voltage readings appear lower than expected despite recent charging, check for these two conditions:
- Voltage Depression: When a battery operates continuously in shallow, mid-range cycles without full saturation charges, its operational plateau can drift downward by 10 to 30 mV per cell. To identify and reverse this, see How to Identify “Lazy Battery” Syndrome (Voltage Depression).
- Current Shunt Drift: Because voltage alone cannot track LFP mid-range capacity, modern systems rely on current-measuring shunts. If an internal shunt drifts or loses calibration, the reported capacity will detach from physical voltage reality. For details on how shunts measure energy flow, review The Role of the Shunt in Measuring Charging Current and How Modern BMS Systems Estimate SoC (Voltage vs. Coulomb Counting).