Operation of Li-Pol batteries at -20 ℃

It is a well-known fact that batteries “freeze” in the cold, lose a significant part of their capacity or are not functional at all. In fact, this is not entirely true.

I assembled a test bench and conducted a series of experiments to check how Li-Pol batteries actually work in the cold. At the same time, I tested a Li-Pol battery from a Russian manufacturer.

Content

How it's done

The Li-Pol battery LP105080-5000 from a Russian manufacturer was selected as a test sample. NETHER, the production of which is located in Kazan.

Battery parameters from official documentation:

• capacity 5000mAh

• discharge voltage 2.5 V, charge 4.2 V

• standard charging current 1 A, fast 2.5 A

• discharge current standard 1 A, fast 2.5 A

• resistance ≤ 40 mOhm

• discharge temperature -40 ℃ to +65 ℃

• dimensions 10 x 50 x 80 mm

• weight 99 g

There are doubts regarding the declared maximum discharge current and internal resistance. But more about that later..

For testing, I assembled a stand from a laptop, a thermometer and an electronic load.

Electronic load ZKETECH EBC-A40L, thermometer UNI-T UT325. Both devices are connected to the laptop via USB. The battery is connected to the electronic load with Kelvin clamps.

The results of the measurements were published in Google Sheets.
You can copy the data and work with it in any convenient environment. For example, build your own graphs in Excel or models in Python and MatchLab.

Testing methodology

Total 6 tests:

1. discharge current 1 A temperature +26 ℃

2. discharge current 1 A temperature -20 ℃

3. discharge current 2.5 A temperature +26 ℃

4. discharge current 2.5 A temperature -20 ℃

5. discharge current 5 A temperature +26 ℃

6. discharge current 5A temperature -20 ℃

Freezing conditions
The cooling temperature of -20 ℃ is chosen due to technical limitations. The freezer does not allow the temperature to be lower than -20 ℃. Both the freezer and the battery itself are cooled to -20 ℃. The thermocouple is attached to the surface of the battery. The battery is in the freezer throughout the measurement process.

About capacity in mAh and energy in Wh

Anticipating comments about “capacity should be measured in W⋅h”, I’ll get ahead of myself and suggest reading the paragraph below and studying the contents of the spoiler.

Capacity is measured in A⋅h, energy in W⋅h.
Capacity Not measured in W⋅h. From W⋅h it is forbidden get A⋅h or vice versa. W⋅h is not an improved version of A⋅h. Energy in W⋅h is not better than capacity in A⋅h. This is different physical quantities.

Below in the spoiler I provide excerpts from GOST, dictionaries and peer-reviewed literature.
Please take a look. For your convenience, even the pages in the books are indicated.

Discharge with current 1 A

Discharge current 1 A, battery temperature +26 ℃.
Discharge current 1 A, battery temperature -20 ℃.

Test results

discharge current 1 A temperature +26 ℃
capacity 4683 mAh
energy 17.73 Wh

discharge current 1 A temperature -20 ℃
capacity 4341 mAh
energy 15.13 Wh

The difference in capacity was only 7%, and the difference in energy was 15%. If you look at the temperature graphs, the reason for the significant difference in the energy given off will become clear. At room temperature, the battery heated up by only 3.3 ℃, and in the freezer by 7.9 ℃. That is, a significant part of the battery energy was spent on heating the battery.

Energy losses “in the cold” are noticeable, but not fatal.

Discharge current 2.5 A

Discharge current 2.5 A, battery temperature +26 ℃.
Discharge current 2.5 A, battery temperature -20 ℃.

Test results

discharge current 2.5 A temperature +26 ℃
capacity 4791 mAh
energy 17.75 Wh

discharge current 2.5 A temperature -20 ℃
capacity 4610 mAh
energy 15.83 Wh

The result is not much different from a discharge with a current of 1 A.

5 A current discharge and the “self-heating effect”

At the very beginning, I mentioned that the declared resistance and discharge current are questionable. The maximum discharge current is 2.5 mA (0.5C) and resistance ≤ 40 mOhm. Although the nominal discharge current for a battery of this type is usually 5 A (1C). The actual measured resistance is only 23 mOhm. Theoretically, the battery should not overheat at 5 A. Overheating calculation below.

Discharge current 5 A, battery temperature +26 ℃.
Discharge current 5 A, battery temperature -20 ℃.

Test results

discharge current 5 A temperature +26 ℃
capacity 4769 mAh
energy 16.98 Wh

discharge current 5 A temperature
capacity 4777 mAh
energy 16.08 Wh

Unlike the two previous tests, the difference between the discharge at +26 ℃ and -20 ℃ is insignificant in the 5 A test. Although the capacity usually drops at high currents. This result is due to the “self-heating effect”. High current contributes to increased heat generation, which leads to rapid heating of the battery. After 7 minutes, the battery temperature rises from -20 ℃ to -4 ℃. After 15 minutes, the temperature is close to 0 ℃. In simple terms, the battery “heats” and defrosts itself. After “self-heating”, the battery's behavior is little different from the operation of a battery at room temperature.

Heating Mathematics. How to predict Li-Pol battery overheating.

It is quite easy to find out to what temperature the battery will heat up at a given current. To do this, you need to know the battery resistance, heat capacity and weight.
Heat capacity is approximately 0.9 J/(g⋅ºС). Resistance and weight can be taken from the documentation or measured independently.

The calculation of the time it takes for the battery to heat up to a given temperature is made using the formula:

• Δt — the temperature to which it needs to be heated
  battery temperature is 26 ºС, the maximum permissible temperature is 65 ºС, therefore, the battery is allowed to heat up by no more than 36 ºС.

• C — specific heat capacity
  The specific heat capacity of the battery is about 0.9 J/(g⋅ºС).

• I — current
  in our case 5 A

• Rin – internal resistance
  actually measured 23 mOhm

Substituting the values ​​into the formula, we get:

Calculations show that it will take 6714 s to heat up to the critical temperature of 65 ºС. At the same time, a full discharge of a 5000 mAh battery will take 3600 s. That is, the battery will discharge faster than it reaches the critical temperature.

Let's see how math compares with practice

The graph shows that the battery discharges in 3440 sec. During this time, it heats up by 20 ºС.

Let's calculate from the same forum how long it will take for the battery temperature to increase by 20 ºС with a resistance of 23 mOhm and a current of 5A. Mathematics gives the following figures:

Estimated time 3130 s. The calculation result is quite close to the measured 3440 s. The actual heating time is slightly longer than the estimated one, since the formula does not take into account the heat loss of the battery through radiation, convection and thermal conductivity of the table.

Why does the battery lose energy?

The drop in battery voltage at low temperatures is due to the increase in internal resistance. This also explains the decrease in energy given to the load and increased heating.

Usually, the internal resistance of a battery is understood as complex resistance, also known as impedance, which includes reactive and active resistance. It is usually measured by a harmonic signal with a frequency of 1 kHz.

However, if you check the battery resistance in the generally accepted way, it turns out that at -20 ℃ the resistance almost does not change. When cooling from +26 ℃ to -20 ℃, the resistance increased from 23 to 26 mOhm.

Measuring impedance may be useful for measuring the resistance of generator windings or the parameters of an acoustic system, but with electrochemical current sources this method does not work quite right.

There is a good article with practical research on ResearchGate and interesting review material on ScienceDirect.

For example, in the article “Effects of temperature on the ohmic internal resistance and energy loss of Lithium-ion batteries under millisecond pulse discharge» on the site ResearchGate describes an experiment in which only ohmic resistancein principle without taking into account the polarization resistance.

What's the bottom line?

Frozen batteries do indeed deliver less energy than batteries at room temperature. However, the difference does not have to be dramatic.

What solutions?

Different battery models behave differently in the cold, so the most obvious solution lies in the area of ​​choosing the element base, that is, the correct selection of batteries suitable for the expected operating conditions.

Another method is to use the “self-heating effect” of the battery. This effect allows you to reach the optimum temperature as quickly as possible. But this solution is already in the field of designing the product itself. By monitoring the temperature and voltage, you can implement a starting mode that will ensure the battery operates in the optimum mode.