Battery voltage on a LiFePO4 bank is the most misread number in off-grid solar. A homeowner on Silvercreek Parkway in Guelph, Wellington County installed a 200Ah LFP battery bank in the summer of 2024. He connected his charge controller display and used the voltage reading as his state-of-charge reference throughout the fall, the same way he had managed his previous lead acid bank. His display showed 13.2V one evening, which would have indicated approximately 80% state of charge on a lead acid battery. He had lights, a DC fridge, and a Starlink dish running. At 2 AM the BMS cut out and everything went dark.
When he checked in the morning, the controller logs showed his bank had been at approximately 15% state of charge when it showed 13.2V, he had been 85% through his usable capacity with no warning from the battery voltage reading. The flat discharge curve is the fundamental property of LFP chemistry. A 12V nominal LFP bank rests at approximately 13.6V at 100% state of charge, approximately 13.2V at 80%, and still approximately 13.0V at 20% state of charge. That is a total battery voltage range of approximately 0.4V across 80% of capacity. By contrast, a lead acid battery drops from 12.7V at full charge to 11.9V at 50% state of charge, a 0.8V spread over the same discharge depth.
After the 2 AM shutdown I walked the Guelph homeowner through the installation of a Victron SmartShunt. The shunt tracks every amp flowing in and out of the battery bank, integrates the current over time, and displays a genuine percentage state of charge accurate to within 1 to 2% in normal operating conditions. His display now shows a number that means something, not a battery voltage reading that looks identical whether he has 90% remaining or 20% remaining.
He has not had an unexpected shutdown since. Understanding why battery voltage is unreliable on LFP is the first diagnostic step. Installing a shunt is the solution. See our Ontario solar sizing guide to confirm your system capacity before relying on any monitoring tool.
Why battery voltage lies on LFP: the flat discharge curve explained
The LFP discharge curve is nearly flat from full to nearly empty. A 12V nominal LFP bank rests at approximately 13.6V at 100% state of charge. It rests at approximately 13.3V at 90%, approximately 13.2V at 80%, approximately 13.1V at 50%, and approximately 13.0V all the way down to 20% state of charge. That total battery voltage span of 0.4V across 80% of usable capacity is what makes the voltage reading nearly useless as a fuel gauge. A basic voltmeter on an LFP bank cannot tell you whether you have 80% or 20% remaining, both read approximately 13.1 to 13.2V at rest.
| State of charge | LFP resting voltage (12V nominal) | Lead acid resting voltage (12V nominal) |
|---|---|---|
| 100% | 13.6V | 12.7V |
| 90% | 13.3V | 12.5V |
| 80% | 13.2V | 12.4V |
| 70% | 13.2V | 12.2V |
| 60% | 13.1V | 12.1V |
| 50% | 13.1V | 11.9V |
| 30% | 13.0V | 11.7V |
| 20% | 13.0V | 11.6V |
| 10% | 12.9V | 11.5V |
Every lead acid state-of-charge chart in every service manual and wall poster was built for sloping-curve chemistry where each voltage reading maps to a specific capacity level. Applying it to LFP is like reading an analogue fuel gauge calibrated for carburettor pressure on a fuel-injected engine, the gauge was designed for a different physical relationship. On LFP, 13.2V is not 80% state of charge. It is anywhere between 90% and 20% depending on load, temperature, and how recently the bank was charged. See our LFP vs AGM chemistry guide for the full electrochemical comparison between these two chemistries.
Battery voltage vs state of charge: the lead acid chart that broke everyone’s LFP setup
The comparison table above confirms the core problem. Lead acid has a 0.8V span from 100% to 50% state of charge, giving a voltmeter approximately 0.016V per percentage point of capacity. That is enough resolution to make useful estimates with a basic meter. LFP has a 0.4V span from 100% to 10% state of charge, giving a voltmeter approximately 0.004V per percentage point, far below the accuracy of any basic voltmeter reading.
When the Guelph homeowner saw 13.2V on his battery voltage display, his best guess would have been 80% state of charge based on the lead acid mental model. The LFP chemistry was at 15%. The gap between what the chart implied and what the coulomb counter knew was 65 percentage points.
A lead acid owner sees 12.1V and knows they are at approximately 60% state of charge from simple table lookup. An LFP owner sees 13.1V and cannot determine whether they are at 80%, 50%, or 25% without a coulomb counter integrated over time. The battery voltage reading on LFP requires a calibrated shunt to convert into useful state-of-charge information. Relying on voltage alone for LFP management is the most common reason Ontario off-grid owners experience unexpected BMS cutoffs, not faulty batteries, not undersized systems, not controller errors. The measurement tool is wrong for the chemistry. See our solar battery lifespan guide to understand how proper monitoring connects to long-term battery health.
Surface charge: why voltage drops after you unplug the charger
Surface charge is the elevated battery voltage that appears immediately after charging stops, before the charge energy redistributes from the surface layer of cells into the bulk material. An LFP bank just removed from absorb or float charge reads approximately 13.8 to 14.4V. After 2 to 4 hours at rest with no load connected, it settles to approximately 13.3 to 13.4V, the true resting voltage. The 0.8 to 1.0V drop is surface charge dissipation, not capacity loss. Zero ampere-hours leave the battery during this process.
A camper on Britannia Road in Milton, Halton County purchased a 100Ah LFP battery in spring 2025 and charged it fully before the first trip, confirming 14.2V at charger completion. When he unplugged shore power at the campsite the display showed 14.4V. Twenty minutes later it showed 13.8V. By the time camp was set up approximately 90 minutes after arrival it read 13.4V. He called the retailer convinced the battery was draining.
The retailer explained surface charge dissipation: the voltage was settling as charge redistributed from cell surfaces into bulk material, with no actual capacity leaving the bank. After 4 hours at rest with no load the battery stabilised at 13.3V, correct for approximately 95% state of charge. The battery performed without issue for the full 2025 season. The surface charge test is simple: wait 4 hours with no load, confirm voltage stabilises and holds steady, confirm the bank is healthy. If voltage continues dropping below 13.0V on a 12V LFP bank with no load attached, check for phantom loads from always-on inverters before concluding battery failure.
Voltage sag under load: the microwave that isn’t killing your battery
Voltage sag is the drop in battery voltage under high current draw, caused by the battery’s internal resistance. LFP internal resistance is approximately 0.02 to 0.05 ohms at normal operating temperature. A 1,200W load on a 12V system draws approximately 100A. At 100A through 0.03 ohms of internal resistance, the voltage drop is approximately 0.3V. The battery voltage display falls from 13.1V to approximately 12.7 to 12.8V the moment the load activates. Remove the load and the display bounces back to 13.1V. This is Ohm’s law applied to an electrochemical source, it is not a sign of battery failure or low state of charge.
The voltage sag confuses new LFP owners because 12.7V on a lead acid chart means approximately 100% discharged. On LFP, 12.7V under a 100A load means the bank is healthy and 80% full. A microwave or electric kettle running at full power causes this exact scenario: battery voltage drops instantly from 13.1V to 12.5 to 12.7V on the display, new owner panics, load turns off, voltage bounces back, battery is fine.
The correct diagnostic tool is the Victron SmartShunt state-of-charge percentage, which remains stable during sag because it is tracking coulombs, not battery voltage. It shows 75% before, during, and after the microwave runs, correctly reflecting that the 0.05 kWh draw from a 2.4 kWh bank represents approximately 2% of capacity, not 50% as the voltage drop falsely implies.
Coulomb counting: why a shunt is the only real fuel gauge
A coulomb counter tracks every amp flowing in and out of the battery bank, integrates the current over time, and calculates state of charge from a known starting point. The Victron SmartShunt uses a 500-microohm shunt resistor. At 100A current, the voltage across the shunt is 100A × 0.0005 ohms = 0.05V = 50mV. The SmartShunt reads this 50mV and calculates the current as 50mV divided by 500 microohms = 100A exactly. It accumulates these measurements over time, Ah out reduces the SoC percentage, Ah in increases it, starting from a user-confirmed 100% baseline after a full charge cycle. The result is a genuine state-of-charge percentage that is independent of the flat battery voltage curve.
The shunt has one important maintenance requirement for accuracy: periodic re-synchronisation at a full charge endpoint. If the battery is never fully recharged to a defined endpoint, the coulomb counter drifts slightly with each cycle. On a well-designed Ontario off-grid system the battery reaches 100% state of charge on most clear summer days, re-syncing the shunt automatically via the charge controller’s absorption endpoint. On a winter system cycling at partial state of charge for extended periods, a manual re-sync at full charge every 2 to 4 weeks maintains accuracy.
The shunt is not perfect, but it is 10 to 20 times more accurate than reading battery voltage on an LFP bank. For installation details see our battery monitor shunt installation guide.
NEC and CEC: battery monitoring and shunt installation code for Ontario
NEC 690 governs solar PV installations including battery monitoring equipment. A shunt monitor wired in line with the battery’s main negative conductor is part of the battery system’s instrumentation. NEC 690.71 requires that the battery system include a means to monitor state of charge or voltage, and a properly installed shunt monitor satisfies this requirement for LFP systems where battery voltage alone is insufficient for accurate state-of-charge measurement.
The shunt must be installed on the main negative conductor between the battery and all loads including the charge controller. Any current path not passing through the shunt will not be counted in the measurement, creating state-of-charge error proportional to the uncounted current. Contact the NFPA at nfpa.org for current NEC 690 battery monitoring requirements.
CEC Section 64 governs battery installations in Ontario off-grid systems. A battery monitoring shunt installed as part of a new battery bank installation is covered under the same ESA permit as the battery bank and requires no separate permit application. A shunt retrofit added to an existing permitted battery bank is a minor modification that typically does not require a permit amendment if the shunt is installed on the existing negative conductor without adding new wiring circuits or changing the system configuration.
The shunt installation must be documented and accessible for inspection by the ESA inspector. Contact the Electrical Safety Authority Ontario at esasafe.com for permit requirements specific to your installation before retrofitting monitoring equipment to a previously permitted battery bank in Ontario.
Pro Tip: The fastest way to confirm your LFP bank is in the danger zone without a shunt is the load test. Connect a known load that draws approximately 10 to 15A, a small fan, a 150W light, or a DC appliance you can measure. Let it run for exactly 1 hour. If the bank has more than 100Ah of capacity, a 10A draw for 1 hour removes only 10Ah, which is 10% of a 100Ah bank. Check the battery voltage before and after the 1-hour test with the load removed and let it rest for 15 minutes. If the voltage drops more than 0.15V over the test period, the bank is already below 30% state of charge and the BMS is approaching cutoff. This test does not replace a shunt, it is a one-time check to determine whether the situation is urgent enough to require emergency action before the shunt arrives. The Guelph Silvercreek Parkway homeowner could have run this test after his first unexplained evening dimming and caught the low-state problem before the 2 AM shutdown. A shunt eliminates the need for any of this manual testing permanently.
The battery voltage verdict: three Ontario owner profiles
- Ontario LFP owner who is currently using battery voltage to manage state of charge: install a Victron SmartShunt before the next weekend away from the system. The Guelph Silvercreek Parkway case shows exactly what happens without one: a bank reading 13.2V appears 80% full on a lead acid mental model while sitting at 15% actual capacity, followed by a 2 AM BMS cutoff with everything downstream going dark simultaneously. Budget approximately $50 to $80 for the SmartShunt and one evening for the installation. The cost of the next unexpected shutdown, food spoilage in a DC fridge, frozen pipes in an off-grid cabin, a missed alarm, exceeds both figures. Battery voltage on LFP is not a management tool. It is a rough indicator that the bank exists and has some charge remaining, nothing more.
- Ontario LFP owner who just disconnected their charger and watched the battery voltage drop 0.8V in 2 hours: the battery is not draining and is not defective. Surface charge dissipation is a universal property of LFP chemistry that occurs regardless of brand, age, or quality of cells. The Milton Britannia Road camper experienced 14.4V dropping to 13.3V over 4 hours with zero load connected and confirmed a completely healthy bank. The surface charge test is: disconnect all loads, wait 4 hours, confirm battery voltage stabilises and holds steady above 13.2V on a 12V bank. If it stabilises, the battery is healthy. If it continues dropping below 13.0V with no load connected, investigate phantom loads from inverters, BMS communication circuits, or accessories left on before concluding battery failure.
- Ontario owner who sees battery voltage drop sharply the moment a high-current load activates: this is Ohm’s law, not battery damage or low state of charge. At 100A draw on a 12V bank, a 0.3V voltage sag is mathematically expected from the battery’s internal resistance of approximately 0.02 to 0.05 ohms. The correct diagnostic is the SmartShunt state-of-charge percentage under load, not the battery voltage reading. If the SmartShunt SoC is above 30% when the load is running, the bank is healthy and the battery voltage reading is the problem. Install the shunt, remove the voltmeter from the dashboard, and stop reading a number that was designed for a different chemistry.
Frequently Asked Questions
Q: Why does my LFP battery voltage stay nearly the same from 90% to 20% state of charge?
A: The flat discharge curve is a fundamental property of lithium iron phosphate cell chemistry. LFP cells maintain a nearly constant electrochemical potential across most of their discharge depth, resulting in only approximately 0.4V of battery voltage change from 90% state of charge down to 10% state of charge on a 12V nominal bank. This is chemically different from lead acid, which shows a 0.8V change over the same discharge depth.
The flat curve makes LFP longer-lived under deep discharge conditions, but it makes battery voltage nearly useless as a state-of-charge indicator. The correct tool is a coulomb-counting shunt such as the Victron SmartShunt, which tracks actual Ah flowing in and out rather than measuring the voltage that the flat LFP curve makes ambiguous.
Q: Why does my battery voltage drop sharply right after I unplug the charger?
A: Surface charge dissipation. During the final absorption phase of charging, electrons accumulate at the surface layer of cells at a higher concentration than in the bulk material. When charging stops, this surface charge slowly redistributes into the bulk, and the battery voltage settles from the surface-elevated level to the true resting level. For LFP this typically means a drop from approximately 14.2 to 14.4V immediately after charging to approximately 13.3 to 13.4V after 2 to 4 hours at rest.
The drop represents zero actual capacity leaving the bank. To confirm: disconnect all loads, wait 4 hours, and check whether the battery voltage stabilises at a steady resting value. If it does, the battery is healthy and the voltage drop was surface charge dissipation.
Q: Is a voltage drop under high loads a sign my battery is failing?
A: No. Voltage sag under high current draw is Ohm’s law applied to a battery’s internal resistance, not a sign of battery failure or low state of charge. LFP internal resistance is approximately 0.02 to 0.05 ohms. At 100A draw, a 1,200W load on a 12V system, the battery voltage drops by approximately 0.3V due to that resistance. Remove the load and the battery voltage returns to its resting level immediately.
The correct way to evaluate battery health under load is the SmartShunt state-of-charge percentage, which does not change due to voltage sag. If the SoC percentage is stable and above 30% while the load runs, the battery is healthy and the voltage reading is the misleading variable, not the battery.
This build is engineered within the 48V DC Safety Ceiling. Diagnostic logic is based on 20+ years of technical service experience. All structural and electrical installations must be verified by a Licensed Professional and comply with your Local AHJ.
This article contains affiliate links. If you purchase through these links, I earn a small commission at no extra cost to you.