Hybrid Electric Vehicle (HEV)

Governments are asking the public to reduce fuel consumption and lower pollution. They do this without imposing a change in driving habits and the HEV fits the bill. Japan is leading in adapting the HEV because of high fuel costs and environmental concerns.

The purpose of the HEV is to conserve fuel without sacrificing performance, and the HEV achieves this by using one or several electric motor to assist the ICE during acceleration and to harness kinetic energy when braking. The ICE turns off at traffic lights and the electric motor propels the car through slow-moving traffic. On full power, both the ICE and electric motor engage for optimal acceleration.

The HEV uses a mechanical powertrain to transfer power from the ICE to the wheels. In this respect, the HEV resembles an ordinary vehicle with a crankshaft and a clutch, also known as parallel configuration. Fuel savings are achieved by the use of a smaller ICE that is tuned for maximum fuel efficiency rather than high torque. Toyota claims a thermal efficiency of 40 percent for the new Prius. Peppy driving is accredited to the electric motor as this propulsion system delivers far better torque than a sluggish ICE of the same horsepower. Figure 1 illustrates the different modes of an electrified powertrain in in an HEV.

Figure 1: Basic function or an electrified powertrain in an HEV.
Battery power is only used for short durations. The HEV battery seldom encounters full charge-discharge cycles that are common in the electric vehicle.
^{Source: RWTH Aachen University, Germany}

Most batteries for HEVs are guaranteed for 8 years. To meet this long service life, the cells are optimized for longevity rather than high specific energy as with consumer products. The battery maker achieves this in part by using a thicker and more durable separator. To reduce stress, the battery operates at 30–80 percent state-of-charge (SoC), or roughly 3.5–4.0V/cell for Li-ion, rather than the customary 3.0–4.20V/cell.

HEV batteries operate momentarily and share similarity with a starter battery by applying short power bursts for acceleration rather than long, continuous discharges as with the EV. Rarely will an HEV battery discharge to a low 20 percent state-of-charge (SoC). Under normal use, a parallel HEV consumes less than 2 percent of the available battery capacity per mile (1.6km). Capacity fade goes unnoticed, and an HEV battery still works well with less than half the original capacity.

Figure 2 shows the battery capacity of six hybrid cars at a 256,000km (160,000 miles). The test was done by the US Department of Energy’s FreedomCAR and Vehicle Technologies Program (FCVT) in 2006 according to SAE J1634 practices and it included the Honda Civic, Honda Insight and Toyota Prius.

Figure 2: End-of-life battery capacity of HEVs. At 256,000km (160,000 miles), the two Honda Civic vehicles had 68% capacity, the Insight had 85% and the Prius had 39%. The capacity fade did not affect the fuel efficiency by much.
^{Source: FreedomCAR and Vehicle Technologies Program}

The hybrid battery of the two Honda Civic vehicles had 68 percent remaining capacity; the Insight had 85 percent and the Prius 39 percent. Even with lower capacity at the end of life, the fuel efficiency was not severely affected. The Insight showed a 1.2mpg (0.12L/km) decrease in fuel economy during the test, while the Prius reduced the fuel efficiency by 3.2mpg (0.33L/km). Air-conditioning was off in both cases.

Stringent battery demands are needed for hybrid trucks with a gross vehicle weight of 33 tons (73,000 lb). The battery must be able to continuously charge and discharge at 4C, deliver 10kW (200hp) for up to 10 minutes, operate at –20°C to 40°C (–4°F to 104°F) and deliver 5 years of service. Supercapacitors would provide the required durability but high cost and low energy density are against this choice. Lead acid has good discharge characteristics but it is slow to charge. Li-ion, especially LTO, would be a good choice but high power draw requires active cooling. Second generation NiMH is being tested; the rugged NiCd may also be tried.

Paradox of the hybrid vehicle

As good as a hybrid may be, the car is not without ironies. At a conference addressing advanced automotive batteries, an HEV opponent argued with an HEV maker that a diesel car offers better fuel economy than a hybrid. Being a good salesman, the HEV maker flatly denied the claim. Perhaps both are right. In city driving, the HEV clearly delivers better fuel-efficiency while diesel consumes less on the highway. Combining both would provide the best solution, but the high cost of a diesel-hybrid solution might not pay back with low fuel prices, although such vehicles are available in Europe.

High-end HEVs come with a full-sized ICE of 250hp and an electrical motor of 150–400hp in total. Such vehicles will surely find buyers, especially if the government assists with grants for being “green.” It’s unfortunate that consumers who walk, cycle or take public transportation won’t get such handouts. Common sense reminds us to conserve energy by driving less, or using smaller vehicles when driving is necessary.

Wolfgang Hatz, the then head of powertrain for Volkswagen Group, said that hybrid technology is a very expensive way to save a small amount of fuel and states that Volkswagen only makes hybrids because of political pressure. He supports diesel as the most energy-efficient motor, especially on highways.

Volkswagen may have a solution — the 1-Liter Car (Figure 3). It is called the 1-Liter Car because the concept vehicle burns only one liter of fuel per 100km. To prove the concept, the then VW chairman Dr. Ferdinand Piëch drove the car from their headquarters in Wolfsburg to Hamburg for a shareholders meeting. The average consumption was just 0.89 liters per 100km (317mpg).

Aerodynamics and weight help to achieve the low fuel consumption. While a typical car has a drag coefficient of 0.30, the 1-Liter Car is only 0.16. Carbon fiber and a magnesium frame reduce the weight to 290kg (640lb). The one-cylinder diesel engine generates 8.5hp (6.3kW), and the 6.5-litre (1.43-gallon) fuel tank has a range of 650 kilometers (400 miles). The average fuel consumption is 0.99 liter per 100km (238mpg).

Although the 1-Liter Car did not go into production, VW demonstrated that fossil fuel could be stretched should the cost rise or should frivolous consumption create unsustainably high pollution levels. Point-to-point personal transportation could be made possible with a light carrier that weighs only 290kg, a weight that is less than the 540kg Tesla S battery. Rather than consuming 150–250Wh per kilometer, as with an electric vehicle, the 1-Liter Car would only use about 40Wh/km. Even though it burns fossil fuel, the environmental impact would be less than an EV propelled with electricity, which is mainly produced by fossil fuel.

Plug-in Hybrid Electric Vehicle (PHEV)

Most PHEVs use a fully electrified powertrain in a series configuration with no mechanical linkage from ICE to wheels. The system runs solely on the electric motor for propulsion, and the ICE only engages when the batteries get low to supply electricity for the electric motor and to charge the battery. The driving range of a fully charged battery is about 50km (30 miles).

The PHEV is ideal for commuting and doing errands. No gasoline is consumed when driving on batteries and the highways are tax-free. However, there will be an increase in the electrical utility bill to charge the batteries at home.

Unlike the parallel HEV that relies on the battery for only brief moments, the PHEV battery is in charge depletion mode, meaning that the battery must work harder than on an HEV. This adds to battery stress and reduces longevity. While a capacity drop to 39 percent will affect the performance of the Toyota Prius HEV only marginally, such a loss would reduce the electric driving range of a PHEV from 50km to 20km (30 to 12 miles).

The Chevy Volt carries a 16kWh Li-ion battery that weighs 181kg (400 lb) and powers a 149hp (111kW) electric motor. The temperature of the prismatic cells is kept at 20–25C (68–77F) during charging and driving. An 115VAC outlet fills the battery in 8 hours; a 230VAC reduces the charging time to 3 hours. The driving range is 64km (40 miles) before the 1.4-liter four-cylinder ICE kicks in to activate the 53kW AC generator that powers the electric motors.

Economics

As good as the PHEV sounds, the long-term savings may be smaller than expected, especially if a battery replacement is needed during the life of the car. Battery aging is an issue that car makers avoid mentioning in fear of turning buyers away. A motorist used to driving ICE cars expects ample power at hot and cold temperatures and minimal performance degradation with age. The battery cannot match this fully, and the owner will need to tolerate a decrease in driving range during the winter, as well as accept a small reduction in delivered mileage with each advancing year due to battery aging.

Modern cars do more than provide transportation; they also include amenities for safety, comfort and pleasure. The most basic of these are the headlights and windshield wipers. Buyers also want cabin heat and air-conditioning, services that are taken for granted in a vehicle that burns fossil fuel. Heat is a by-product in the ICE that must be generated with battery power in a PHEV, but the larger concern is air-conditioning, which draws 3–5kW of power. Comforts might need to be provided more sparingly when running on a battery.

Many PHEV buyers value the environmental benefit and the pleasure of driving a quiet vehicle powered by electricity. This has a large buyer appeal because electric propulsion is more natural than that of an ICE. Drivers must adapt to the new lifestyle of charging the vehicle at night when electricity is cheap and then driving measured distances. Users of these cars will also appreciate new charging stations at workplaces and shopping malls.

Understanding the importance of low conductivity

Capacity alone is of limited use if the pack cannot deliver the stored energy effectively; a battery also needs low internal resistance. Measured in milliohms (mΩ), resistance is the gatekeeper of the battery; the lower the resistance, the less restriction the pack encounters. This is especially important in heavy loads such as power tools and electric powertrains. High resistance causes the battery to heat up and the voltage to drop under load, triggering an early shutdown. Figure 1 illustrates a battery with low internal resistance in the form of a free-flowing tap against a battery with elevated resistance in which the tap is restricted.

Low resistance, delivers high current on demand; battery stays cool.

High resistance, current is restricted, voltage drops on load; battery heats up.

Figure 1: Effects of internal battery resistance. A battery with low internal resistance delivers high current on demand. High resistance causes the battery to heat up and the voltage to drop. The equipment cuts off, leaving energy behind.

Lead acid has a very low internal resistance and the battery responds well to high current bursts that last for a few seconds. Due to inherent sluggishness, however, lead acid does not perform well on a sustained high current discharge; the battery soon gets tired and needs a rest to recover. Some sluggishness is apparent in all batteries at different degrees but it is especially pronounced with lead acid. This hints that power delivery is not based on internal resistance alone but also on the responsiveness of the chemistry, as well as temperature. In this respect, nickel- and lithium-based technologies are more responsive than lead acid.

Sulfation and grid corrosion are the main contributors to the rise of the internal resistance with lead acid. Temperature also affects the resistance; heat lowers it and cold raises it. Heating the battery will momentarily lower the internal resistance to provide extra runtime. This, however, does not restore the battery and will add momentary stress.

Crystalline formation, also known as “memory,” contributes to the internal resistance in nickel-based batteries. This can often be reversed with deep-cycling. The internal resistance of Li-ion also increases with use and aging but improvements have been made with electrolyte additives to keep the buildup of films on the electrodes under control. With all batteries, SoC affects the internal resistance. Li-ion has higher resistance at full charge and at end of discharge with a big flat low resistance area in the middle.

Alkaline, carbon-zinc and most primary batteries have a relatively high internal resistance, and this limits their use to low-current applications such as flashlights, remote controls, portable entertainment devices and kitchen clocks. As these batteries deplete, the resistance increases further. This explains the relative short runtime when using ordinary alkaline cells in digital cameras.

Two methods are used to read the internal resistance of a battery: Direct current (DC) by measuring the voltage drop at a given current, and alternating current (AC), which takes reactance into account. When measuring a reactive device such as a battery, the resistance values vary greatly between the DC and AC test methods, but neither reading is right or wrong. The DC reading looks at pure resistance (R) and provides true results for a DC load such as a heating element. The AC reading includes reactive components and provides impedance (Z). Impedance provides realistic results on a digital load such as a mobile phone or an inductive motor.

Figure 2 illustrates the internal resistance of an 18650 Li-ion cell when exposed to 1,000 full cycles at 40ºC (104ºF). The AC readings in the green frame do not reflect the true resistive state of a battery; DC method provides more reliable performance data with loading.

Figure 2: Rise of internal resistances of 18650 Li-ion cell measured with AC and DC methods when cycled.
AC resistance readings in green frame stay low; DC method gives true state.
^{Source: Technische Universität München (TUM)}

Pack Resistance

The internal resistance of a battery does not consist of the cells alone but also includes the interconnection, fuses, protection circuits and wiring. In most cases these peripherals more than double the internal resistance and can falsify rapid-test methods. Typical readings of a single cell pack for a mobile phone and a multi-cell battery for a power tool are shown below.

Internal Resistance of a Mobile Phone Battery

Internal Resistance of a Power Pack for Power Tools

^{Source: Siemens AG (2015, München)}

Figures 3, 4 and 5 reflect the runtime of three batteries with similar Ah and capacities but different internal resistance when discharged at 1C, 2C and 3C. The graphs demonstrate the importance of maintaining low internal resistance, especially at higher discharge currents. The NiCd test battery comes in at 155mΩ, NiMH has 778mΩ and Li-ion has 320mΩ. These are typical resistive readings on aged but still functional batteries. That demonstrates the relationship of capacity, internal resistance and self-discharge.)

Figure 3: GSM discharge pulses at 1, 2, and 3C with resulting talk-time
The capacity of the NiCd battery is 113%; the internal resistance is 155mΩ. 7.2V pack.

Figure 4: GSM discharge pulses at 1, 2, and 3C with resulting talk-time
The capacity of the NiMH battery is 94%, the internal resistance is 778mΩ. 7.2V pack

Figure 5: GSM discharge pulses at 1, 2, and 3C with resulting talk-time
The capacity of the Li-ion battery is 107%; the internal resistance is 320mΩ. 3.6V pack

Notes: The tests were done when early mobile phones were powered by NiCd, NiMH and Li-ion. Li-ion and NiMH have since improved.

The maximum GSM draws is 2.5A, representing 3C from an 800mAh pack, or three times the rated current.

Discover what a battery needs to get going and maintain a long life.

In many ways, a battery behaves like a human being. It senses the kindness given and delivers on the care given. It is as if the battery has feelings and returns on the benevolence bestowed. But there are exceptions, as any parent raising a family will know; and the generosity conferred may not always deliver the anticipated returns.

To become a good custodian, you must understand the basic needs of a battery, a subject that is not taught in school. This section teaches what to do when the battery is new, how to feed it the right “food” and what to do when putting the pack aside for a while. Chapter 7 also looks into restrictions when traveling with batteries by air and how to dispose of them when their useful life has passed.

Just as a person’s life expectancy cannot be predicted at birth, neither can we date stamp a battery. Some packs live to a great old age while others die young. Incorrect charging, harsh discharge loads and exposure to heat are the battery’s worst enemies. Although there are ways to protect a battery, the ideal situation is not always attainable. This chapter discusses how to get the most from our batteries.

Priming a New Battery

Not all rechargeable batteries deliver the rated capacity when new, and they require formatting. While this applies to most battery systems, manufacturers of lithium-ion batteries disagree. They say that Li-ion is ready at birth and does not need priming. Although this may be true, users have reported some capacity gains by cycling after a long storage.

“What’s the difference between formatting and priming?” people ask. Both address capacities that are not optimized and can be improved with cycling. Formatting completes the fabrication process that occurs naturally during use when the battery is being cycled. A typical example is lead- and nickel-based batteries that improve with usage until fully formatted. Priming, on the other hand, is a conditioning cycle that is applied as a service to improve battery performance during usage or after prolonged storage. Priming relates mainly to nickel-based batteries.

Lead Acid

Formatting a lead acid battery occurs by applying a charge, followed by a discharge and recharge. This is done at the factory and is completed in the field as part of regular use. Experts advise not to strain a new battery by giving it heavy duty discharges at first but gradually working it in with moderate discharges, like an athlete trains for weight lifting or long-distance running. This, however, may not be possible with a starter battery in a vehicle and other uses. Lead acid typically reaches the full capacity potential after 50 to 100 cycles. Figure 1 illustrates the lifespan of lead acid.

Deep-cycle batteries are at about 85 percent when new and will increase to 100 percent, or close to full capacity, when fully formatted. There are some outliers that are as low as 65 percent when tested with a battery analyzer. The question is asked, “Will these low-performers recover and stand up to their stronger brothers when formatted?” A seasoned battery expert said that “these batteries will improve somewhat but they are the first to fail.”

The function of a starter battery lies in delivering high load currents to crank the engine, and this attribute is present from the beginning without the need to format and prime. To the surprise of many motorists, the capacity of a starter battery can fade to 30 percent and still crank the engine; however, a further drop may get the driver stranded one morning. See also BU-904: How to Measure Capacity)

Nickel-based

Manufacturers advise to trickle charge a nickel-based battery for 16–24 hours when new and after a long storage. This allows the cells to adjust to each other and to bring them to an equal charge level. A slow charge also helps to redistribute the electrolyte to eliminate dry spots on the separator that might have developed by gravitation.

Nickel-based batteries are not always fully formatted when leaving the factory. Applying several charge/discharge cycles through normal use or with a battery analyzer completes the formatting process. The number of cycles required to attain full capacity differs between cell manufacturers. Quality cells perform to specification after 5–7 cycles, while lower-cost alternatives may need 50 or more cycles to reach acceptable capacity levels.

Lack of formatting causes a problem when the user expects a new battery to work at full capacity out of the box. Organizations using batteries for mission-critical applications should verify the performance through a discharge/charge cycle as part of quality control. The “prime” program of automated battery analyzers (Cadex) applies as many cycles as needed to attain full capacity.

Cycling also restores lost capacity when a nickel-based battery has been stored for a few months. Storage time, state-of-charge and temperature under which the battery is stored govern the ease of recovery. The longer the storage and the warmer the temperature, the more cycles will be required to regain full capacity. Battery analyzers help in the priming functions and assure that the desired capacity has been achieved.

Lithium-ion

Some battery users insist that a passivation layer develops on the cathode of a lithium-ion cell after storage. Also known as interfacial protective film (IPF), this layer is said to restrict ion flow, cause an increase in internal resistance and in the worst case, lead to lithium plating. Charging, and more effectively cycling, is known to dissolve the layer and some battery users claim to have gained extra runtime after the second or third cycle on a smartphone, albeit by a small amount.

Scientists do not fully understand the nature of this layer, and the few published resources on this subject only speculate that performance restoration with cycling is connected to the removal of the passivation layer. Some scientists outright deny the existence of the IPF, saying that the idea is highly speculative and inconsistent with existing studies. Whatever the outcome on the passivation of Li-ion may be, there is no parallel to the “memory” effect with NiCd batteries that require periodic cycling to prevent capacity loss. The symptoms may appear similar but the mechanics are different. Nor can the effect be compared to sulfation of lead acid batteries.

A well-known layer that builds up on the anode is the solid electrolyte solid electrolyte interface (SEI). SEI is an electrical insulation but has sufficient ionic conductivity to allow the battery to function normally. While the SEI layer lowers the capacity, it also protects the battery. Without SEI, Li-ion might not get the longevity that it has. (See BU-307: How does Electrolyte Work?)

The SEI layer develops as part of a formation process and manufacturers take great care to do this right, as a batched job can cause permanent capacity loss and a rise in internal resistance. The process includes several cycles, float charges at elevated temperatures and rest periods that can take many weeks to complete. This formation period also provides quality control and assists in cell matching, as well as observing self-discharge by measuring the cell voltage after a rest. High self-discharge hints to impurity as part of a potential manufacturing defect.

Electrolyte oxidation (EO) also occurs on the cathode. This causes a permanent capacity loss and increases the internal resistance. No remedy exists to remove the layer once formed but electrolyte additives lessen the impact. Keeping Li-ion at a voltage above 4.10V/cell while at an elevated temperature promotes electrolyte oxidation. Field observation shows that the combination of heat and high voltage can stress Li-ion more than harsh cycling.

Lithium-ion is a very clean system that does not need additional priming once it leaves the factory, nor does it require the level of maintenance that nickel-based batteries do. Additional formatting makes little difference because the maximum capacity is available right from the beginning, (the exception may be a small capacity gain after a long storage). A full discharge does not improve the capacity once the battery has faded — a low capacity signals the end of life. A discharge/charge may calibrate a “smart” battery but this does little to improve the chemical battery. (See BU-601: Inner Working of a Smart Battery.) Instructions recommending charging a new Li-ion for 8 hours are written off as “old school,” a left-over from the old nickel battery days.

Non-rechargeable Lithium

Primary lithium batteries, such as lithium-thionyl chloride (LTC), benefit from passivation in storage. Passivation is a thin layer that forms as part of a reaction between the electrolyte, the lithium anode and the carbon-based cathode. (Note that the anode of a primary lithium battery is lithium and the cathode is graphite, the reverse of Li-ion.)

Without this layer, most lithium batteries could not function because the lithium would cause a rapid self-discharge and degrade the battery quickly. Battery scientists even say that the battery would explode without the formation of lithium chloride layers and that the passivation layer is responsible for the battery’s existence and the ability to store for 10 years.

Temperature and state-of-charge promote the buildup of the passivation layer. A fully charged LTC is harder to depassivate after long storage than one that was kept at a low charge. While LTC should be stored at cool temperatures, depassivation works better when warm as the increased thermal conductivity and mobility of the ions helps in the process.

The passivation layer causes a voltage delay when first applying a load to the battery, and Figure 2 illustrates the drop and recovery with batteries affected by different passivation levels. Battery A demonstrates a minimal voltage drop while Battery C needs time to recover.

Figure 2: Voltage behavior when applying a load to a passivated battery.
Battery A has mild passivation, B takes longer to restore, and C is affected the most.
^{Courtesy EE Times}

LTC in devices drawing very low current, such as a sensor for a road toll or metering, may develop a passivation layer that can lead to malfunction, and heat promotes such growth. This can often be solved by adding a large capacitor in parallel with the battery. The battery that has developed a high internal resistance is still capable of charging the capacitor to deliver the occasional high pulses; the standby time in between is devoted to recharging the capacitor.

To assist in sulfation prevention during storage, some lithium batteries are shipped with a 36kΩ resistor to serve as a parasitic load. The steady low discharge current prevents the layer from growing too thick, but this will reduce the storage life. After 2-year storage with the 36kΩ resistor, the batteries are said to still have 90 percent capacity. Another remedy is attaching a device that applies periodic discharge pulses during storage.

Not all primary lithium batteries recover when installed in a device and when a load is applied. The current may be too low to reverse the passivation. It is also possible that the equipment rejects a passivated battery as being low state-of-charge or defective. Many of these batteries can be prepared with a battery analyzer (Cadex) by applying a controlled load. The analyzer then verifies proper function before engaging the battery in the field.

The required discharge current for depassivation is a C-rate of 1C to 3C (1 to 3 times of the rated capacity). The cell voltage must recover to 3.2V when applying the load; the service time is typically 20 seconds. The process can be repeated but it should take no longer than 5 minutes. With a load of 1C, the voltage of a correctly functioning cell should stay above 3.0V. A drop to below 2.7V means end-of-life. (See BU-106: Primary Batteries)

These lithium-metal batteries have high lithium content and must follow more stringent shipping requirements than Li-ion of the same Ah. (See BU-704a: Shipping Lithium-based Batteries by air) Because of the high specific energy, special care must be taken in handling these cells.

Explore the limitations when operating a battery at adverse temperatures and learn how to minimize the effects.

Like humans, batteries function best at room temperature. Warming a dying battery in a mobile phone or flashlight in our jeans might provide additional runtime due to improved electrochemical reaction. This is likely also the reason why manufacturers prefer to specify batteries at a toasty 27°C (80°F). Operating a battery at elevated temperatures improves performance but prolonged exposure will shorten life.

As all drivers in cold countries know, a warm battery cranks the car engine better than a cold one. Cold temperature increases the internal resistance and lowers the capacity. A battery that provides 100 percent capacity at 27°C (80°F) will typically deliver only 50 percent at –18°C (0°F). The momentary capacity-decrease differs with battery chemistry.

The dry solid polymer battery requires a temperature of 60–100°C (140–212°F) to promote ion flow and become conductive. This type of battery has found a niche market for stationary power applications in hot climates where heat serves as a catalyst rather than a disadvantage. Built-in heating elements keep the battery operational at all times. High battery cost and safety concerns have limited the application of this system. The more common lithium-polymer uses gelled electrolyte to enhance conductivity.

All batteries achieve optimum service life if used at 20°C (68°F) or slightly below. If, for example, a battery operates at 30°C (86°F) instead of a more moderate lower room temperature, the cycle life is reduced by 20 percent. At 40°C (104°F), the loss jumps to a whopping 40 percent, and if charged and discharged at 45°C (113°F), the cycle life is only half of what can be expected if used at 20°C (68°F).
The performance of all batteries drops drastically at low temperatures; however, the elevated internal resistance will cause some warming effect by efficiency loss caused by voltage drop when applying a load current. At –20°C (–4°F) most batteries are at about 50 percent performance level. Although NiCd can go down to –40°C (–40°F), the permissible discharge is only 0.2C (5-hour rate). Specialty Li-ion can operate to a temperature of –40°C but only at a reduced discharge rate; charging at this temperature is out of the question. With lead acid there is the danger of the electrolyte freezing, which can crack the enclosure. Lead acid freezes quicker with a low charge when the specific gravity is more like water than when fully charged.

Figure 1 illustrates the discharge voltage of an 18650 Li-ion under various temperatures. A 3A discharge of a 2.8Ah cell represents a C-rate of 1.07C. The reduced capacity at low temperature only applies while the cell is in that condition and will recover in room temperature.

Figure 1: Discharge voltage of an 18650 Li-ion cell at 3A and various temperatures.
Cell type: Panasonic NRC18650PD, 2.8Ah nominal, LiNiCoAlO2 (NCA)
^{Source: Technische Universität München (TUM)}

Matched cells with identical capacities play an important role when discharging at low temperature and under heavy load. Since the cells in a battery pack can never be perfectly matched, a negative voltage potential can occur across a weaker cell in a multi-cell pack if the discharge is allowed to continue beyond a safe cut-off point. Known as cell reversal, the weak cell gets stressed to the point of developing a permanent electrical short. The larger the cell-count, the greater is the likelihood of cell-reversal under load. Over-discharge at a low temperature and heavy load is a large contributor to battery failure of cordless power tools. 

The driving range of an electric vehicle between charges is calculated at ambient temperature. EV drivers are being made aware that frigid temperature reduces the available mileage. This loss is not only caused by heating the cabin electrically but by the inherent slowing of the battery’s electrochemical reaction, which reduces the capacity while cold.