Cold Batteries: Why Charging Needs Its Own Limits

A battery can have energy available for discharge while accepting much less power in the opposite direction. At low temperature, charging a graphite-based lithium-ion cell involves a competition: lithium can enter the graphite structure, or it can deposit as metal on the electrode. The second route can return some charge later and still cause lasting damage. This makes cold-weather charging a separate operating constraint, rather than a mirror of discharge power.

Charging asks the negative electrode to do a different job

During normal charging, lithium ions enter the graphite negative electrode through intercalation: lithium is stored within the layered structure. During discharge, lithium leaves that structure. Charging also drives the graphite potential closer to the potential at which lithium metal can form. Cold conditions and higher current can increase polarization enough for metal deposition to compete with intercalation. This electrode potential is measured relative to lithium; it is not the terminal voltage displayed for a complete cell or cabinet. Chen et al., introduction and Figures 2–3.

A low-temperature discharge result therefore cannot establish an equally permissible charging rate. The direction of current changes the electrochemical process and its limiting conditions. Discharge still has temperature-dependent power and energy limits, but it does not ask the graphite electrode to accommodate incoming lithium in the same way. The distinction matters after a cold night, when solar generation or an off-peak tariff may create a strong reason to recharge quickly.

A cold cell may store charge outside the intended structure

Zinth and colleagues studied a commercial 1.95 Ah cell with a nickel–manganese–cobalt oxide (NMC) positive electrode and a graphite negative electrode at −20 °C using neutron diffraction. They followed changes in lithium-containing graphite phases during slower C/30 and faster C/5 charging, followed by either a 20-hour rest or immediate discharge. Those C-rates were defined against the cell’s nominal capacity. The experiment could track graphite lithiation without opening the cell; metallic lithium itself was difficult to detect directly. Zinth et al., pp. 153–156.

After C/5 charging, the graphite was less fully lithiated than after the slower charge. During rest, it continued to become more lithiated even though no additional charging current entered the cell. The authors interpreted this as deposited lithium moving into graphite. Immediate discharge produced a different sequence: charge left the cell before the graphite phases began their usual change. The authors attributed 374 mAh, equivalent to about 19% of the cell’s nominal room-temperature capacity, to oxidation of plated lithium. This was an estimate from the combined diffraction and discharge observations, not a direct measurement of a metal deposit. Zinth et al., Figures 7 and 10, pp. 156–158.

The practical implication is subtle: charge can pass into and out of a cell without all of it following the intended graphite-storage route. A subsequent capacity result alone cannot prove that the preceding cold charge was benign. Part of the deposited lithium may remain available; part can become inactive or contribute to unwanted surface reactions. Recoverable charge and absence of degradation are different outcomes.

The study was a narrowly controlled cell experiment, not a test of a stationary LFP cabinet. Its slow reference cycle also experienced a voltage disturbance that temporarily raised the current, and the authors found evidence consistent with some plating even after that reference charge. C/30 must therefore not be read as a demonstrated universal safe rate. The reported fractions belong to this cell and protocol. Zinth et al., pp. 153, 156 and 158.

Why cold and high current can create a similar problem

A 2025 study by Chen and colleagues examined graphite more directly. A separate reference electrode allowed the researchers to distinguish the graphite potential from the potential of the lithium-metal counter electrode. That separation matters: a voltage feature measured across two electrodes can include a contribution from the counter electrode, rather than describe graphite alone. The experiment combined controlled temperature and current changes with analysis of graphite’s charging features. Chen et al., methods and Figure 2.

At a low current density of 0.1 mA/cm², approximately C/24 for their graphite loading, the familiar graphite intercalation features remained apparent down to −20 °C. At −30 °C and −40 °C, the graphite potential fell much earlier toward the lithium-plating region and those features were strongly reduced. Raising current at 20 °C produced qualitatively similar changes. The main curves used graphite without prior formation cycling, which normally develops the solid electrolyte interphase (SEI) on the electrode. Separate tests with a pre-formed SEI showed additional differences at −30 and −40 °C. These results connect cold charging and fast charging through the electrode’s ability to accommodate the imposed current; they do not establish −20 °C as a safe charging threshold for other cells. Chen et al., Figures 3–4 and 6.

The authors proposed an explanation involving lithium-ion adsorption, movement between available sites and electron transfer. In their framework, ions can accumulate at graphite sites when the subsequent steps cannot keep pace with the imposed charging current. The graphite potential then shifts toward conditions that favor metal deposition. Their temperature analysis distinguished the multistep behavior of intercalation from the behavior associated with plating. This is a mechanistic interpretation supported by their measurements, rather than a complete model of every commercial electrode.

The distinction adds more than the familiar statement that cold batteries have higher resistance. Improving electrolyte transport alone need not remove limitations at the graphite interface. At the same time, the study’s model assumes reaction kinetics dominate over electrolyte mass transport in its tested configuration. Different electrode thicknesses, electrolytes, ageing states and cell designs can change the balance. A laboratory current density cannot be translated directly into a cabinet charging limit. Chen et al., kinetic analysis and variable-current discussion.

LFP describes the positive electrode, not every charging limit

Lithium iron phosphate describes the positive electrode chemistry. An LFP cell paired with a graphite negative electrode can still encounter graphite-side lithium plating. Petzl’s 2015 doctoral research studied this explicitly in commercial LFP/graphite cells. Its low-temperature ageing experiments at −22 °C varied charge rate and state-of-charge range, combining electrochemical measurements with examination of opened cells. Higher current and a wider upper charge range increased degradation under those test conditions. Petzl, doctoral thesis, Sections 5.2–5.3.

The LFP result connects the electrode mechanism to a chemistry widely used in stationary storage, but it does not supply a universal temperature cutoff or lifetime prediction. The relevant charging behavior still depends on the particular negative electrode, electrolyte, cell construction, state of charge and control limits. Naming the positive electrode chemistry does not resolve those details.

For a solar-storage system, timing matters as well as energy

The system consequence is that nominal kilowatt-hours do not describe how quickly a cold battery can absorb incoming energy. A solar-storage installation may have available capacity while its allowed charging power is temporarily low. If solar production exceeds the combined site load and permitted battery charging power, the remainder follows the site’s available export or curtailment path. A larger nameplate energy rating, by itself, does not establish a higher permitted charging power under those cold conditions.

A system-level implication is that limited charging power can make a recharge deadline harder to meet. Energy restored before the next dispatch depends on the charging power permitted throughout the intervening period, together with conversion and auxiliary losses. An off-peak window can close before the target stored energy is reached even when there is enough nominal capacity to hold it.

Temperature history adds another time scale. Outdoor air warming after sunrise does not mean the cells have immediately reached the same temperature. A system that includes battery heating may use energy and time to reach its allowed charging conditions; the effect depends on the installed hardware and controls. Liquid cooling alone does not establish that a cabinet includes a heater, nor does an ambient-temperature range specify a cold-start charging sequence. The charging envelope is therefore a relationship among cell temperature, state of charge and permissible current, not just a single outdoor-temperature number.

What the PowerFac-261 specifications do—and do not—say

The HUA-X PowerFac-261 is a liquid-cooled commercial and industrial system with 125 kW rated AC power, 261 kWh nominal battery energy and a 400 V AC connection. Its published product information lists LFP cells and an operating-temperature range of −20 to 55 °C, with output derating above 45 °C. These values describe the model’s stated ratings; they do not define a temperature-and-SOC charging map. PowerFac-261 procurement brief.

For a PowerFac-261 project, available charging power over the site’s temperature range shapes how much energy can be restored within the recharge window. Model-specific charging limits, the supplied thermal-control configuration and any warm-up behavior determine that answer. Discuss those PowerFac-261 operating conditions with HUA-X.

Cold-weather storage therefore has two distinct capabilities: delivering energy already stored and replenishing it within the time available. A useful system specification connects both directions to their operating conditions. The research explains why a battery’s willingness to discharge cannot stand in for its ability to accept the next charge.