Don’t murder your batteries, tips for winter storage of LFP batteries

Editor’s note: Please join me in welcoming Allen Jones to Panbo. Allen has been an active participant on numerous electrical forums and Facebook groups. Allen also runs the YouTube channel MV Intrigue – Trawler DIY. Additionally, Allen worked in technical support for Epoch Batteries and moderated their facebook user support groups.

In recent years LiFePO4 (LFP) batteries in boats have morphed from a slightly mysterious luxury to commonplace. The perception has gone from an expensive and risky modification to a widely accepted, game-changing, and comfort enhancing upgrade. As such there is great excitement when it’s “your turn” to upgrade your boat. Marinas and boating forums are full of stories from the recently converted about their newly installed LFP system and the resulting improved and simplified boating. Those still using legacy lead for house power may be feeling left behind.  At the same time the prices of converting to LFP have dropped significantly.

I have been able to share these revolutionary experiences with hundreds of people while working with the tech support department of a popular battery manufacturer. I have had a front row seat to witness the joys of making the change and improving their boating experience. LFP truly is a game changer and in this case the cliche is apt.

However, there is a rarely discussed potential dark cloud hanging over LFP batteries. Once you understand the cause it is as obvious as it is devastating. Over discharging cuts short the lives of many LFP batteries. Most boaters know that overly discharged batteries are bad. While many boaters are careful not to allow over-discharge of their new LFP bank, many lose track of their batteries during storage periods, and some may fail to track them at all. The great majority of us have killed a car or boat battery at some point. But most of our experience is with cheaper and potentially more forgiving batteries

Excitement over the new technology and shared stories about the amazing performance of LFP adds to the mystique. All the promise and performance aside, you can murder an LFP battery just as quickly as a lead acid battery. In some cases, maybe even faster.

There are many misunderstandings about LFP conversions. Many of these misunderstandings stem from the differences between LFP and lead acid battery chemistry and physics. An LFP graph of the voltage vs SOC contains most of the clues to these mysteries. It’s important to untangle and understand what kills an LFP battery. With this understanding, users can manage their batteries, so they live a long life and provide reliable service year after year.

As many know, nearly all LFP batteries contain a battery management system or BMS. The BMS serves to keep the battery within a “safe operating envelope” or “SOE”. It does so by monitoring parameters such as voltage, current and temperatures. When one of those parameters reaches an unsafe threshold, the BMS intercedes and isolates the battery. For MOSFET based BMS’s the charge and discharge MOSFETS are individually controlled.  For batteries that use contactors to control the flow of power, the battery is entirely isolated from the DC system.

MOSFETs are essentially electronic switches that manage the flow of electricity in or out of the battery as needed. If the voltage gets too high the charge MOSFET will shut off to keep the cell/pack voltage within the SOE. When the voltage gets too low it shuts off the discharge MOSFET preventing cell damage. With these fancy features it is understandable you may assume a quality LFP battery can safely manage itself. If the BMS is protecting the cells, how can you end up with an overly discharged battery to the point of damage?

Unfortunately, when boats come out of storage, many are finding completely dead batteries or banks. The majority are battery banks that won’t power up, won’t take a charge and are essentially non-functional. How could this possibly happen to these miracle batteries?

To understand the details, we need some battery basics. All batteries have some amount of self-discharge.  LFP cell specifications typically quote 1-3% per month. This number is generally accurate for the mid-range state of charge between 5% and 99% SOC. But the 1-3% range doesn’t apply when the cells reach what is known as “the lower knee”.

As seen in the voltage vs SOC chart below, once an LFP battery nears depletion, the drop in voltage is precipitous. The bottom row of the chart represents amp hours (Ah).  Ah is a unit of capacity; however, it is a unit made up of two components, time and current flow. Taken another way, the precipitous drop in voltage is graphed over a time component on the bottom row. Self-discharge of LFP cells must be factored into this precipitous drop when cell voltage is in the lower knee ranges.

As a result, you can see that a self-discharge in the mid-range of SOC over the course of what might be a month will hardly be noticed regarding voltage. But a month or even a few weeks with cell voltage in the lower knee is a recipe for disaster. Cold temperatures often encountered during winter storage hasten this trajectory towards the lower knee voltages.

The final factor is battery electronics energy consumption. Things like Bluetooth modules, the BMS itself, indicator lights, and displays all use some power. Even small internal draws will add up when calculated over months. In addition, I have noted significant variation in self-consumption between battery brands and even models of a single brand. BMS selection and additional features such as plug in monitors seem to explain much of the spread. Some batteries can have an astounding amount of self-consumption while others have nearly none.

The BMS may or may not properly register cell self-discharge and parasitic consumption. As a result, SOC calculation in the battery app or monitor may become significantly inaccurate over long idle periods. It’s not uncommon to see a battery registering a full or nearly full SOC with a voltage reading indicating near complete discharge.

The battery BMS calculates SOC using an internal shunt to measure power in and out of the battery and then calculates power usage. Most self-contained batteries shunts lack sufficient resolution to capture the very small currents being used in self-consumption and have no means to capture actual self-discharge. In addition, some BMSs will shut off when the battery is powered down or when the battery “goes to sleep” and stops capturing this SOC loss over time.

As a result, you may shut down the battery at 13.3 volts and 75% SOC. After several months turned off, it still reads 75%. But, in our hypothetical scenario, voltage has dropped to 12.7 volts, indicating a large loss in capacity not captured by the SOC meter. Not all batteries behave this way but be sure to keep a close eye out for this occurrence. These challenges are not deal breakers but do need to be known by the boat owner so they can keep a keen eye on things. A good rule of thumb is to get used to looking at state of charge percentages and voltages at each check. Always cross check.

The actual cause of battery homicide generally fits into just 2 categories:

1. Batteries that did not have sufficient capacity to be stored for the length of time they were stored. Especially in very cold temperatures.
2. Batteries and DC systems that were not prepared for the anticipated inactive period properly prior to winter storage.

Capacity sufficient to avoid cause number one above is determined by the rates of cell self-discharge and battery self-consumption. Typically, self-consumption is the larger culprit. I have seen batteries with self-consumption as high as 0.25 amps/hour. That doesn’t sound like much but consider a 300ah battery. That calculates to around 50 days of storage before the battery exhausts 300ah of energy just to support itself. And having 2 or 4 of these very same 300ah batteries does not increase that 50-day mark since each battery is consuming its own power. At the 50 day point the battery voltage is squarely in the lower knee and the BMS will shut off the discharge MOSFETs to save itself. This usually happens around 2.5 volts per cell or 10.0 to 10.5 volts of the pack. But shutting off the discharge MOSFETs may or may not help to slow the death spiral.

Each passing hour with voltage in the lower knee drops capacity and voltage at an increasing rate. Eventually, it falls so low that the BMS cannot be powered any longer. At this point the battery is already in danger of being unrecoverable. If it is on the cusp of BMS voltage, applying a charge at this time may or may not power up the BMS and allow charging to commence. Past this stage, internal intervention is required to directly charge the cells which requires opening the battery (if possible) and charging the cells directly at contact points prior to the BMS. In this state, if the battery is not discovered and rescued quickly, the voltage will continue to fall until the cells themselves are damaged. The battery is in a precarious state. You have days or a few weeks at most to discover it and take immediate action to charge and rescue your LFP batteries.

It is unfortunate, but many times that rescue never comes. When a boat owner returns to a dead boat with an unresponsive DC system, the natural assumption is a failure of their newly installed LFP batteries. All 2,3,4 or more of them, at once! Although LFP batteries do fail from time to time, I assure you they don’t fail 4 at a time without help. This cascade of events typically culminates in a call to tech support.  What happens next varies based on many factors. Are the batteries serviceable so that the top may be removed and cells probed? Is the customer honest with the description of events? The battery manufacturers policy and even the sophistication of support in troubleshooting the event will affect the outcome.

I can tell you that many times batteries may get replaced under warranty for damage that probably shouldn’t have been covered. In other circumstances the warranty is denied completely based on details and information gathered. Several times I led the customer, via troubleshooting, to the realization that he/she essentially killed their batteries, and that it is not a warrantable condition. It is these conversations that led me to write this article. It’s a tough conversation to have and it is far too common.

To avoid being caught in such a situation I have refined my opinion on LFP-battery winter storage to just a few cardinal rules. In my opinion there are only three acceptable practices for LFP winter storage.

1. If shore power is available leave the battery pack on charge throughout the winter. Set a storage voltage of 13-13.2 volts and leave the charger on. If you want an increased time cushion in case of power loss, use the standard float voltage of 13.5 volts or the manufacturer’s recommended float. Check on the system often to ensure power remains on and the batteries are being maintained. Bluetooth or the Victron VRM remote monitoring is very useful here. Even a 120-volt AC outlet with a small charger will be sufficient. If the battery has self-heating features, you must check on the battery during the initial cold snaps in which the self-heating and charging should have engaged to ensure proper heating and charging. If you anticipate temperatures low enough to initiate internal heaters, ensure there is sufficient amperage available from the charger to power all heaters in the system simultaneously as some heaters in large batteries require 1 to 10 amps to bring the temperature up before a charge will be allowed. In some scenarios that could be a significant amount of power, and more than a small charger can handle.
2. Remove the batteries. Take them home and put them in a warm place and on charge using a quality charger such as the Victron IP65 or IP22 with storage mode engaged or charge to a mid-level SOC and shut them down. Even in this scenario, you should still check them often.
3. The least desirable option in my opinion is to charge the batteries to nearly full and secure all possible loads, preferably with battery disconnect switches. Making a checklist for this scenario is particularly useful. It would be wise to check on the batteries weekly for the first month to evaluate power losses and adjust the interval of the visits as needed. Check after prolonged periods of cold weather and charge as required. Routine monitoring is still a must. Sometimes this is not possible with boats that are wrapped and fully winterized. In that case another option is mandatory.

Regardless of what you choose, walking away from the boat for 3-6 months is a bad option. Protect your investment!