One of the most common failures after a switch to LiFePO4 goes like this: the engine is running, the alternator is charging, and suddenly the chartplotter, autopilot and radio go dark. Sometimes they never come back on. Sometimes the alternator stops charging. The cause can be the BMS disconnecting the battery during charging.
ABYC E-13, the standard for lithium batteries, devotes a separate appendix to this and explicitly requires the installation to be ready for it. In this article I explain where these voltage spikes come from and how to protect against them. You can read more about the standards in ABYC vs ISO 13297: comparing the standards.
If you have conventional lead-acid batteries and a 1–2–BOTH switch, have a look at my earlier article too, Battery switch with the engine running: will you damage the alternator this way?. That one is about turning the switch through OFF. This one is about the case where the BMS, not you, disconnects the battery.
Every lithium installation must have a BMS (ABYC E-13, clause 13.6.1.1). The BMS makes sure the battery works within the safe operating limits set by the manufacturer: voltage, current and temperature. When one of those limits is exceeded, the BMS first tries a gentler response and, if that does not help, cuts the battery off.
Typical reasons for a disconnect during charging:
The standard requires the BMS to disconnect the battery independently of chargers, regulators and inverters (clause 13.7.1.1). In other words, the BMS does not wait for the alternator to stop charging.
The standard also notes two things few people know:
The battery manufacturer must also document the disconnects that may occur in normal operation (clause 13.8.1.7). It is worth reading that before designing the installation.
An alternator, a shore-power charger or an MPPT solar controller pushes current into the battery, and the battery absorbs it and holds the voltage of the whole system steady. When the BMS suddenly disconnects the battery:
The most dangerous sources are those with high inductance. E-13 names high-output alternators, chargers and inverter/chargers specifically. An alternator is essentially a coil with a rotating magnetic field. When the load is suddenly removed, the energy stored in its windings has to discharge somewhere.
A voltage spike like that can damage:
For comparison, the European standard ISO 13297 (clause 5.4) requires DC equipment to work from 75% to 133% of the nominal voltage, which is 9 V to 16 V on a 12 V system. An alternator spike can go far beyond that range.
The E-13 appendix describes a situation where a spike damages the installation so badly that every electrical system and the propulsion stop working, a dead-ship condition. This is not theory: a boat with a single lithium bank that starts the engine and powers the navigation can be left without engine, lights and plotter after a BMS disconnect.
The standard stresses that voltage spikes can also happen with lead-acid batteries, for example when you turn off the battery switch while the alternator is charging, or when a fuse blows or a contactor opens during charging. With LiFePO4, though, it happens far more often, because disconnecting is built into the BMS.
The standard is clear (clause 13.6.10.1): the boat's electrical system must be designed so that a BMS disconnecting the battery does not damage other components.
It also requires charging sources to follow the charging profile recommended by the battery manufacturer (clause 13.6.2.1), and prohibits connecting loads directly to the battery in a way that bypasses the BMS (clause 13.6.1.2), unless the manufacturer explicitly allows it.
The standard's appendix sets out two ways of dealing with voltage spikes.
The BMS talks to the alternator (through an external regulator), the charger and the solar controller:
This is the best solution. It does, however, need equipment that can communicate, for example through the BMS's "allow to charge" output or a communication bus, and above all it has to actually be connected. On many boats I find a BMS with that output wired to nothing.
If the equipment does not communicate, the spike has to be absorbed. The standard lists:
These measures can be built into the BMS, into the charging sources, or fitted as separate devices.
E-13 recommends an alternative power source for critical systems: engine starting, propulsion, navigation lights. It can even be a second lithium battery. In practice this arrangement often works well:
The lead-acid battery on the alternator side also acts as the buffer the standard mentions.
| Question | What the standard says |
|---|---|
| Can the BMS disconnect the battery without warning? | Yes, even routinely while charging (E-13, appendix; 13.8.1.7) |
| Must the installation be designed for it? | Yes, a disconnect must not damage other components (E-13, 13.6.10.1) |
| Most dangerous sources | high-output alternators, chargers, inverter/chargers |
| Protection with communication | the BMS reduces charging or shuts the source down before disconnecting |
| Protection without communication | avalanche diodes, suppressors, grounding, a lead-acid buffer battery |
| Power for critical systems | an alternative source is recommended (E-13) |
| Equipment voltage range | 9–16 V on a 12 V system (ISO 13297, clause 5.4) |
Planning a switch to LiFePO4, or already on lithium and seeing strange electronics failures? I will design or check your installation to ABYC E-13, including alternator protection and backup power. Write or call me, including on WhatsApp: +39 351 384 9089 or +48 791 74 74 54.
Grzegorz Szubert, Yacht Electrical Services, ABYC member
Client area · Online work order · Privacy & cookies · This site is protected by reCAPTCHA and is subject to the Privacy Policy and Terms of Service of Google.