What a BMS does and does not do

Why a lithium battery needs electronics in the first place

A lead-acid battery forgives errors via chemical means. If it is overcharged, it begins to gas: The excess energy is used to decompose water rather than destroy the plates. This costs water and service life, but it acts as a built-in pressure relief valve. Likewise, in a lead-acid battery bank, individual cells rebalance themselves during extended charging because the full cells simply continue to gas while the weak ones are still taking a charge.

Note Some BMS units relay their values externally. A BullTron battery with P-Bus enablement reports the state of charge, cell voltages, and temperatures directly to the PCU control unit — for details, see the article Integrating a BullTron battery directly into the P-Bus.

A lithium iron phosphate cell lacks this mechanism. It takes on charge until it is full, after which the voltage rises very rapidly. There is no chemical secondary path to dissipate the excess energy — only damage. The same applies at the lower end: during deep discharge, the potential at the anode reverses, and the copper collector begins to dissolve. The dissolved copper then deposits during the next charge cycle and can create internal short circuits. This is why the BMS shutdown threshold of approximately 2.5 to 2.8 volts per cell is set well above the range where this occurs.

Then there is the serial connection. A 12-volt LiFePO4 battery consists of four cells of 3.2 volts each in series; larger blocks consist of several such strings in parallel. All cells in a string carry the same current, but they are never exactly identical: manufacturing tolerances, temperature differences within the housing, and slightly different self-discharge rates cause the states of charge to drift apart over months. The battery is full when the strongest cell is full, and empty when the weakest one is empty. Without balancing, the usable capacity shrinks to that of the weakest cell — and that cell is the first to hit overvoltage during every charging cycle.

This is exactly what the battery management system is for: it sees every cell individually, not just the terminal voltage of the block.

The six tasks of a BMS

1. Measure cell voltages individually

Terminal voltage reveals nothing about the state of the individual cells. A 12.8-volt block can consist of four cells at 3.20 volts — or three at 3.05 volts and one at 3.65 volts. From the outside, both look the same, but in the second case, the battery is on the verge of a shutdown. This is why a dedicated measuring lead runs from the BMS to every cell connection.

2. Cell balancing

With passive balancing, the BMS connects a resistor to the most advanced cell and burns off a few hundred milliamperes until the others have caught up. This only works at the top of the charging curve, where cell voltages differ at all, and it takes time — the balancing current is tiny compared to the charging current. Passive balancing is the norm and is perfectly sufficient for onboard systems, provided that the battery bank is regularly fully charged.

With active balancing, charge is shifted from the full cell to the empty cell instead of being burned off. This is more complex and expensive, and is mainly worthwhile for large battery banks that rarely reach the final charging voltage.

Note Anyone who only operates their lithium battery bank between 30 and 80 percent for months at a time may be sparing the cells, but never gives the balancing system a chance to work. A full charge every few weeks is not a leftover requirement from the lead-acid world, but rather the prerequisite for ensuring the cells remain synchronized.

3. Monitor temperature and enable charging

Below approximately 0 °C, lithium deposits metallically on the anode during charging instead of intercalating into the graphite lattice. This process is irreversible, permanently consumes active material, and cannot be detected from the outside — the battery looks healthy and simply has less capacity. The BMS therefore measures cell temperature and blocks the charging enable signal below the cell manufacturer's limit. These limits vary: some manufacturers specify -5 °C, others set +5 °C. Discharging at sub-zero temperatures, however, is permissible.

4. Monitor current

The BMS recognizes both a continuous current limit and a peak current limit. The time constant is important: to prevent inrush currents from anchor winches or inverters from shutting down the system every time they are switched on, the overcurrent shutdown is deliberately slow — typically reacting only after the condition has persisted for several tens of seconds. Remember this figure, as it is the core of the next section.

5. Disconnecting — separately for charging and loads

A BMS that simply shuts everything off during an overvoltage event has a problem: it then also disconnects the consumers, even though discharging would be exactly the right reaction. Good designs therefore have two separate paths — one for charging sources, one for loads. In the event of cell overvoltage, the charge path is disconnected; in the event of cell undervoltage, the load path is disconnected.

6. Communicating with charging equipment

The most elegant solution is to not have to disconnect at all. For this reason, a BMS reports via bus or switching contact to chargers, solar charge controllers, and alternator regulators that they should power down before the shutdown threshold is reached. Crucially: this signal must go to every charging source. If one forgets to wire the charge enable signal when retrofitting a solar controller, it will continue to charge until the BMS performs a hard disconnect.

Typical threshold values

Typical ranges for LiFePO4 onboard batteries. Your cell manufacturer's values take priority — they can be found in the battery data sheet, not the charger manual.

What a BMS expressly does NOT do

This section is the reason for this article. Most planning errors do not occur because someone sets the BMS incorrectly, but because they attribute tasks to it that it simply does not have.

It is not a fuse

A BMS shuts down with a delay during overcurrent because it must not interpret inrush currents as a fault. A short circuit must be disconnected in milliseconds, not seconds. During this time span, the full short-circuit current of the battery bank flows through the line — and because of the very low internal resistance, this is significantly higher in LiFePO4 batteries than in a lead-acid battery bank of the same size. The overcurrent protection fuse near the positive terminal is therefore not a concession to standard regulations, but the only protection that is fast enough. It requires an interrupting rating that matches the short-circuit current of the battery bank; for larger lithium battery banks, this regularly leads to Class T style fuses.

It does not protect the charging sources

It does not replace a charging profile

A BMS does not regulate, it limits. The charger sets the end-of-charge voltage, absorption time, and float voltage. Anyone who connects a lead-acid charger to a lithium battery bank and trusts that the BMS will handle it is running the system permanently at the limit: every charge cycle ends with an emergency shutdown instead of a clean end of charge.

It is not a state-of-charge indicator

The BMS knows cell voltages. The discharge curve of LiFePO4 is almost flat over large sections — between 90 and 20 percent state of charge, the difference is often less than 0.3 volts. No reliable state of charge can be derived from this voltage. For that, you need an amp-hour counter with a shunt in the common negative line, which counts the actual charge that has flowed.

It is not an authorization for maintenance work

An electronically disconnected BMS can switch back on at any time, such as when a cell recovers or a charger is connected. For work on the system, you need a mechanical disconnection — the main battery switch or a removed fuse.

Designs: internal, external, single or dual-channel

With an integrated BMS, the electronics are inside the battery housing. This is convenient and correct for standard cases; the interface to the outside usually consists of a bus connection and a switching contact for charge enable or safety relays. Disadvantage: you are bound to the manufacturer's logic and, in the event of failure, often cannot continue to operate the battery without a workshop.

With an external BMS using individual cell monitoring, the electronics are outside and control one or two contactors. This allows more freedom in system design, but requires that someone understands the threshold values and sets them correctly.

More important than internal or external is the question of single or dual-channel design. A single-channel system has one contactor for everything: in the event of cell overvoltage, it also disconnects the consumers, even though discharging would be the correct reaction; and in the event of deep discharge, it also disconnects the charging sources, meaning the battery bank cannot recover without intervention. Dual-channel designs with separate charge and load paths avoid both. Standards do not explicitly demand this — ABYC E-13 requires a BMS that keeps the battery in its safe operating range and disconnects it otherwise, but makes no specifications regarding separate paths. It is a design choice, not a mandate, and in practice, a good one.

What happens after a deep discharge

If the BMS has disconnected due to cell undervoltage, the battery is invisible to the onboard electrical system — and many chargers will not start because they do not detect any battery voltage. Manufacturers solve this via revival using a low current: the cells are brought to about 2.5 volts with roughly one ampere, followed by a self-test to check if the deep discharge caused any damage. Depending on the condition, this takes anywhere from half an hour to several hours.

The cause is important: A deep discharge during operation means the battery bank is undersized or a consumer is running unnoticed. A deep discharge in winter storage means quiescent current consumers were not disconnected. Both will repeat if only the battery is revived without correcting the underlying cause.

Typical errors

• Trusting the BMS instead of a fuse. Two different time constants, two different components.
• Wiring the charge enable only to the main charger. Solar controllers, alternators, and shore power chargers each need it individually.
• Placing safety-critical consumers behind the BMS. Bilge pumps, anchor lights, and emergency VHF radios belong on a separate supply.
• Never fully charging. Running permanently in the middle state-of-charge window prevents balancing.
• Reading cell voltage as state of charge. The curve is too flat for that.
• Only resetting after a BMS shutdown. The shutdown is the notification, not the fault.
• Forgetting the BMS's quiescent current. Even the electronics themselves consume power, which can drain the battery over months in winter storage.

What you should check in your system

First: Is there an overcurrent fuse between the battery and the distribution board whose interrupting rating matches the short-circuit current of the battery bank? Second: Does every charging source receive the BMS charge enable signal, including retrofitted ones? Third: What happens to your safety-relevant consumers if the BMS disconnects? Fourth: Is there an amp-hour counter with a shunt, or are you inferring state of charge from voltage? And fifth: When was the battery bank last fully charged?

Anyone who can answer these five questions has already eliminated the most common causes of failure in onboard lithium systems.