Battery Thermal Management: Heat Generation, Cooling & TIM

Temperature is the single variable with the greatest influence on battery performance, lifetime, and safety. It affects every electrochemical process inside the cell — reaction kinetics, ion transport, SEI growth rate, lithium plating risk, electrolyte viscosity, and ultimately, thermal runaway. It is the variable the thermal management system exists to control.

The challenge is that batteries are thermally demanding in two opposing ways. Under charge and discharge they generate heat that must be removed — failure to do so causes overheating, accelerated degradation, and at the extreme, thermal runaway. In cold conditions, the cells must be heated before charging can begin safely. And at all times, every cell in the pack must be kept within 2°C of every other cell — because a cell running hotter than its neighbours ages faster, diverges in capacity, and eventually becomes the weakest link that limits the whole pack.

Temperature uniformity matters as much as absolute temperature. A ±5°C gradient across cells in a series string means those cells age at different rates. Over thousands of cycles, this divergence accumulates until cell imbalance becomes a capacity-limiting constraint. Thermal management design for uniformity is as important as thermal management for absolute maximum temperature.

Thermal management is therefore not a secondary system added after the electrical design is complete. It is a primary design constraint that shapes cell selection, module architecture, pack enclosure geometry, and BMS control strategy simultaneously.

Where to Start

New to battery thermal management? This is a suggested reading order through the core material. For the complete list of every page in this section, see the Full Index below.

1. Heat Generation in a Cell – the whole picture: where the thermal load actually comes from.
2. Thermal Management System – the system-level response to that heat.
3. Passive Cooling – the simplest cooling approach.
4. Cooling Plate Design – the most common active cooling approach.
5. Coolants – what actually flows through those cooling plates.
6. Battery Heating – the flip side: warming a pack in cold conditions.
7. Thermal Interface Materials – how heat actually transfers between components.

Full Index

A complete, always up to date index of every page in the Thermal section, organised by hierarchy. Click any heading to expand it.

Why Temperature Matters

Temperature affects battery systems across four interconnected areas:

Performance — cell capacity and available power are both strongly temperature-dependent. At −10°C, a typical NMC cell delivers around 70–80% of its 25°C capacity. DCIR increases 3–5× at −10°C relative to 25°C, dramatically reducing peak power. The thermal management system must ensure the cells are warm enough to perform before power is demanded.

Lifetime — degradation mechanisms are thermally activated. SEI growth rate approximately doubles for every 10°C increase in temperature (Arrhenius kinetics). A pack operating at 35°C rather than 25°C can expect to age 2× faster in calendar life. Conversely, charging at low temperature risks lithium plating, which reduces capacity and can create internal short circuits. The operating temperature window directly determines how long the pack lasts.

Uniformity — cells at different temperatures have different internal resistances and different SoC-voltage relationships. In a series string, the hotter cells deliver more current per unit voltage change, causing them to diverge in SoC from the cooler cells. Over thousands of cycles this divergence accumulates into cell imbalance that reduces usable pack capacity. The ±2°C uniformity target is not arbitrary — it is derived from modelling the rate of imbalance accumulation.

Safety — above approximately 80–130°C (chemistry-dependent), exothermic reactions begin inside the cell that increase temperature further, driving more reactions, in a self-reinforcing cascade: thermal runaway. The thermal management system is the primary line of defence against overtemperature, and the pack structure is designed to contain the consequences if it fails.

Heat Generation in a Cell

Heat is generated in a cell whenever current flows. The total heat generation rate has two components:

Irreversible (Joule) heating: Q_irreversible = I² × R_int

This is always positive (always generates heat) and dominates at high charge/discharge rates. It depends on current squared and internal resistance — doubling the current quadruples the heat generation. DCIR increases with age and at low temperature, so an aged cell in a cold environment generates significantly more heat per unit of current than a new cell at 25°C.

Reversible (entropy) heating: Q_reversible = I × T × (dU/dT)

This term depends on the entropic coefficient of the cell chemistry and can be endothermic (the cell cools) or exothermic (the cell heats) depending on chemistry, SoC, and direction of current. For NMC cells during discharge, the entropy term is exothermic at low SoC and endothermic at high SoC. For LFP cells, the entropy term is relatively small. At low discharge rates where Joule heating is small, the entropy term can dominate — causing unusual temperature behaviour such as cooling during the early phase of discharge.

Total heat: Q_total = I(V − U_oc) + I × T × (dU/dT)

where V is terminal voltage, U_oc is open circuit voltage, T is temperature in Kelvin, and dU/dT is the entropic coefficient.

At high C-rates (fast charge, peak discharge), the Joule heating term dominates and the reversible term can be neglected for a first estimate. At rates below 0.5C, the entropy term must be included for accurate thermal modelling.

→ Heat generation in a cell — full derivation

Specific Heat Capacity

The specific heat capacity of the cell determines how quickly its temperature rises for a given rate of heat generation. A cell with high thermal mass (high Cp and high mass) is more tolerant of transient heat pulses — it absorbs energy without a large temperature rise.

Specific heat capacity values:

LFP’s higher Cp means it absorbs more energy per degree of temperature rise than NMC or NCA — one small contribution to LFP’s better thermal stability, alongside its more thermally stable olivine lattice.

At pack level, the effective Cp of the assembly is lower than the cell Cp alone because the enclosure, busbars, and TIM all contribute mass with lower specific heat capacity than the cell active materials.

First-order temperature rise estimate:

ΔT = Q_pulse / (m_cell × Cp)

For a 60Ah NMC cell (mass 1.0kg, Cp = 1040 J/kg·K) generating 50J during a 10s peak discharge: ΔT = 50 / (1.0 × 1040) = 0.048°C

This confirms that single pulse events rarely cause temperature problems in well-designed packs — it is sustained high-rate operation over minutes that drives temperature above limits.

→ Specific heat capacity of lithium-ion cells

Thermal Conductivity and Temperature Gradients

Heat generated in the core of a cell must be conducted to its outer surface before it can be transferred to the cooling system. The thermal conductivity of the cell’s active layers determines how easily this happens — and it is highly anisotropic.

In-plane thermal conductivity (parallel to the electrode layers) is dominated by the metallic current collectors — aluminium (cathode) and copper (anode) — and is typically 15–160 W/m·K. Heat flows easily along the electrode plane.

Through-plane thermal conductivity (perpendicular to the electrode layers, through the separator and active material) is much lower: typically 0.2–0.8 W/m·K. Heat generated at the centre of a thick cell stack must conduct through many layers of poor thermal conductor to reach the outer surface.

This anisotropy has direct implications for cooling design. Cooling through the face of a pouch cell (through-plane) is limited by the low through-plane conductivity. Cooling at the edge (in-plane) uses the higher conductivity path along the electrode layers — which is why edge cooling is the preferred design for high-performance pouch cell modules.

Temperature gradient from Fourier’s Law:

ΔT = Q × L / (k × A)

where Q is heat flux (W), L is conduction path length (m), k is thermal conductivity (W/m·K), and A is cross-sectional area (m²). Halving the conduction path length (L) or doubling the cross-sectional area (A) halves the temperature gradient for the same heat flux.

Pressure matters. The through-plane thermal conductivity of uncompressed pouch cells degrades by up to 23% over lifetime. With appropriate compression maintained by the module structure, this degradation reduces to approximately 4%. This is one of several reasons why maintaining cell compression pressure throughout pack lifetime is important — thermal conductivity is as much a reason as capacity retention.

→ Thermal conduction in a cell → Thermal conductivity of the active layers → Cell temperature gradient → Thermal conduction and temperature gradients

Temperature Limits

Every cell has a set of temperature limits that define the boundaries of safe and optimal operation. These limits are chemistry and manufacturer-specific — always use the cell manufacturer’s specified limits.

Operating temperature ranges (indicative, lithium-ion):

Why charging below 0°C is dangerous: At low temperatures, the rate of lithium intercalation into graphite slows significantly due to reduced ion diffusion and higher SEI resistance. If the charge rate exceeds the rate at which the anode can accept lithium, metallic lithium plates on the anode surface rather than intercalating. This plated lithium is electrochemically inactive (capacity loss) and can form dendrites — conductive filaments that grow towards the cathode and can eventually penetrate the separator, causing an internal short circuit. The risk is higher at high SoC, high C-rate, and cold temperature — all simultaneously present during fast charging in cold weather.

Why high temperature accelerates degradation: The rate of SEI growth follows Arrhenius kinetics — approximately doubling for every 10°C increase. A cell operating at 40°C calendar-ages approximately 2× faster than the same cell at 30°C. Above 45°C, electrolyte decomposition and CEI growth on the cathode also accelerate significantly.

→ Temperature limits of a battery → Cold temperature charge/discharge → Cell temperature

Cooling System Architectures

The cooling architecture is one of the most consequential design decisions in battery pack development. It determines heat removal capacity, pack mass and volume, cost, manufacturing complexity, and long-term reliability.

Passive Cooling

No external energy input — heat is rejected by conduction through the pack enclosure and natural convection/radiation to the environment. Viable only for low heat generation rates: small packs, low C-rate applications, or cells with inherently low DCIR (LFP at low rates).

Advantages: no pumps, no coolant, no leak risk, simplest design, lowest cost. Limitation: heat removal rate is limited by the temperature difference between cells and ambient. As ambient temperature rises, cooling capacity falls.

→ Passive cooling

Forced Air Cooling

Air is moved over the cells by a fan — either through channels between cells or over the pack exterior. Practical for HEV packs and some small automotive applications where heat generation is moderate.

Advantages: simple, no liquid circuits, low cost. Limitations: air has low thermal conductivity (0.026 W/m·K) and low volumetric heat capacity, giving poor heat transfer coefficients (~25–100 W/m²·K). Achieving uniform cooling across a large cell population is difficult; large temperature gradients are common. Not suitable for fast charging or high continuous power applications.

Liquid Cooling — Cooling Plate

The dominant architecture for automotive EV and large BESS applications. A metal plate (usually aluminium) with internal channels carries a water-glycol coolant mixture in contact with or close to the cell surfaces via a thermal interface material (TIM).

How it works: Coolant enters the plate at one end (typically at the coldest temperature), absorbs heat from the cells, and exits at the other end warmer. The temperature of the coolant rises along the flow path, creating an inherent cell-to-cell temperature gradient — cells near the coolant outlet see warmer coolant and run hotter than cells near the inlet. Careful flow path design (serpentine, parallel, or counter-flow) minimises this gradient.

Heat transfer coefficient: Typically 500–3,000 W/m²·K depending on coolant velocity and channel geometry — significantly better than air cooling.

Key design parameters:

• Coolant inlet temperature: typically 15–25°C for normal operation
• Coolant flow rate: higher flow reduces cell-to-outlet temperature rise but increases pumping power
• Plate material: aluminium for low mass and high thermal conductivity; must be electrically isolated from cells
• TIM selection: gap pad, phase-change material, thermal adhesive, or potting compound — see TIM section below

For the 21700 cylindrical cell format, base cooling (coolant plate under the cell base) provides approximately 12% greater heat flux than side cooling for the same temperature gradient, due to the larger contact area at the base.

→ Cooling plate design → Base vs side cooling — cylindrical cells → Tesla fast charge thermal model

Liquid Cooling — Dielectric Immersion

The cells are fully submerged in a dielectric (electrically non-conductive) liquid. The dielectric is in direct contact with all exposed cell surfaces, eliminating the conduction resistance of the TIM and cooling plate.

Heat transfer coefficients of 500–3,000 W/m²·K (single-phase) to 10,000+ W/m²·K (two-phase boiling) are achievable. This enables significantly higher sustained charge and discharge rates before cell temperature limits are reached.

Advantages: excellent cell-to-cell uniformity (all cells immersed in the same coolant), no TIM, no cooling plate, inherent pressure equalisation, and automatic thermal runaway containment (the dielectric absorbs and dilutes vented gas and heat).

Limitations: dielectric fluids are expensive; the system must be hermetically sealed and the dielectric must be compatible with all cell and pack materials over the pack lifetime; pumping losses are higher than water-glycol systems; fluid replacement or replenishment at service is more complex.

→ Dielectric immersion cooling

Phase Change Material (PCM)

A phase change material (typically a wax or salt hydrate) is placed around or between cells. As the cells heat up, the PCM absorbs thermal energy by melting — the latent heat of fusion absorbs large amounts of energy with minimal temperature rise. When the cells cool, the PCM re-solidifies and is ready for the next event.

PCM is a passive buffering system, not an active cooling system. It handles transient heat spikes — peak discharge pulses, short fast charge events — but cannot reject heat continuously. In most applications, PCM is combined with a low-power liquid or air cooling system: the PCM handles peaks, and the steady-state system recharges the PCM between events.

Active Refrigerant Cooling

A heat pump or vapour compression refrigeration cycle directly cools the coolant circuit or the cells. Enables cell temperatures significantly below ambient — important for fast charging in hot climates where ambient temperature alone would otherwise limit charge rate.

Used in several premium production vehicles and in BESS installations in high-ambient-temperature climates. Adds significant system complexity and energy consumption.

→ Battery cooling options — full comparison

Format-Specific Cooling Considerations

Pouch Cells

Pouch cells present a large flat face but low through-plane thermal conductivity. The preferred cooling approach is edge cooling — a cooling plate contacts the cell at its edges, using the in-plane conductivity of the electrode layers as the primary heat conduction path. This is more effective than face cooling despite using a smaller contact area, because the in-plane conductivity is 10–100× higher.

The cell-to-cooling-plate thermal path must account for: TIM conductivity and thickness, dimensional tolerance of the cell and module, and pressure applied to the cell stack. The cell’s in-plane thermal conductivity allows the heat to spread from the bulk of the electrode to the cooled edge efficiently, provided the electrode geometry is designed to minimise the conduction path length.

→ Pouch cell cooling → Pouch cell design and edge cooling

Cylindrical Cells

Cylindrical cells can be cooled at the base, the side, or the top. In most modern designs, base cooling is preferred for 21700 and 4680 cells — the cooling plate sits under all cells simultaneously, providing a large contact area. The move to tabless electrode designs (Tesla 4680, LG 4695) improves heat spreading to the cell can surface, reducing the thermal gradient between the centre of the jellyroll and the cooled base.

Single-side busbars — connecting both positive and negative terminals on the top of the cell — leave the base entirely free for the cooling plate, eliminating the busbar-to-coolant thermal interference present in double-side designs.

→ Base vs side cooling — cylindrical cells → Single side 21700 busbars

Prismatic Cells

Prismatic cells are typically cooled through their large flat faces — the cooling plate contacts the face of the cell stack, with TIM between cell and plate. The module compression system that applies pressure to the cell faces also ensures consistent TIM contact throughout the cell’s lifetime as the cell swells.

Thermal Interface Materials (TIM)

The thermal interface material fills the gap between the cell surface and the cooling plate, providing a low-resistance thermal path across the dimensional tolerance between the two surfaces. Without TIM, even a small air gap (very low thermal conductivity: 0.026 W/m·K) creates a significant thermal resistance that can dominate the total cell-to-coolant thermal resistance.

TIM types and typical properties:

Selection criteria:

• Thermal conductivity — must be high enough to keep ΔT across TIM within budget
• Thickness — accommodates dimensional variation of cell and cooling plate; must remain effective as cells swell over lifetime
• Electrical isolation — most TIMs are electrically insulating; verify dielectric strength at operating voltage
• Chemical compatibility — with cell case materials, coolant, and enclosure over temperature and lifetime
• Manufacturability — paste/grease requires dispensing; pads can be pre-cut; adhesive requires cure management
• Reworkability — structural adhesives and potting are effectively permanent; gap pads can be removed at service

→ Thermal interface materials → Options for TIM → TIM application pattern → Thermally conductive adhesives

Battery Heating

Thermal management is not only cooling — it also includes heating. In cold climates, cells must be brought to a minimum temperature before charging can begin, and before the full power capability of the pack is available for discharge.

Why heating matters:

• Most lithium-ion cells should not be fast-charged below 5°C and should not be charged at any rate below −10°C (NMC) or 0°C (LFP), due to lithium plating risk
• DCIR increases dramatically at low temperature, severely limiting available power
• Vehicle range at −10°C is 20–40% lower than at 20°C, partly due to reduced battery capacity and partly due to increased cabin heating load

Heating methods:

PTC heating elements — positive temperature coefficient resistive heaters embedded in the module or applied to the cooling plate. Self-limiting: resistance increases with temperature, naturally limiting heat output as the cell warms. Simple and safe, but electrically resistive heating is energy-inefficient.

Heat pump — uses the vehicle’s refrigerant circuit in reverse, extracting heat from the environment and delivering it to the coolant circuit that heats the cells. Coefficient of performance (COP) of 2–4 means significantly less electrical energy consumed than resistive heating for the same heat delivered.

AC pre-conditioning — while plugged into the grid, the vehicle heats (or cools) the battery before departure. This preserves battery energy for driving rather than consuming it for temperature management — a significant range benefit in cold weather.

→ Battery heating

Battery Thermal Management System (BTMS)

The Battery Thermal Management System is the complete system — hardware and control — that regulates battery temperature. It encompasses: the heat generation in the cells, the conduction path through the TIM and cooling plate, the coolant circuit with pump and heat exchanger, the heating elements or heat pump, the temperature sensors, and the BMS control logic that operates all of it.

BTMS control objectives:

1. Keep all cells below maximum operating temperature
2. Keep all cells above minimum operating temperature (for charging)
3. Maintain cell-to-cell temperature uniformity within ±2°C
4. Pre-condition cells for fast charging when grid connection is available
5. Minimise energy consumption of the cooling/heating system

The BMS manages the BTMS by monitoring cell temperatures from the sensor array and varying coolant pump speed, coolant inlet temperature (via the chiller or heater), and heating element power. In some systems, the BMS can also derate charge/discharge current if temperature targets cannot be maintained — trading performance for thermal safety.

Cell temperature sensing placement — sensors placed between cells at the module centre (the hottest location in most flow geometries) and adjacent to the coolant inlet and outlet are most informative. The BMS can only protect what it can measure. Sensors on the module housing or the cooling plate measure coolant temperature, not cell temperature — a common design mistake that results in late detection of cell overtemperature.

→ Thermal management system (BTMS) → Cell temperature sensing → Cell to cell temperature gradient

Thermal Runaway and Propagation

When a cell’s temperature rises above approximately 80–130°C (chemistry-dependent), exothermic decomposition reactions begin inside the cell. These reactions release heat that raises the temperature further, driving more reactions — a self-reinforcing cascade that is self-sustaining once started. This is thermal runaway.

The thermal management system prevents thermal runaway under normal operating conditions. But the pack must also be designed to contain the consequences if a single cell enters thermal runaway — preventing the event from propagating to neighbouring cells and potentially consuming the entire pack.

Thermal propagation is the spread of thermal runaway from one cell to adjacent cells via:

• Conductive heat transfer (hot cell heats adjacent cells by conduction)
• Hot vent gas (the venting cell ejects hot flammable gas)
• Projectile / core ejection (in cylindrical cells, the cell contents can be ejected at high velocity)
• External fire (from burning electrolyte igniting adjacent cells or pack structure)

Pack-level mitigation strategies:

• Cell spacing or barriers — physical separation between cells, or thermal barrier materials (mica, aerogel, ceramic) between cells to limit conductive heat transfer
• Vent paths — designed openings that direct hot gas away from adjacent cells and out of the pack
• Fire suppression — automatic suppression systems within the pack for large-format applications
• Pack enclosure venting — burst discs or venting membranes that relieve pack pressure to prevent catastrophic failure of the enclosure

→ Thermal runaway → Thermal propagation → Mica as a thermal barrier → Fire suppression

Fast Charging Thermal Constraints

Fast charging imposes the most demanding transient thermal load on a battery pack. The peak I²R heat generation during a 3C charge is 9× greater than at 1C. For the same DCIR, the same pack that runs comfortably at 1C may generate heat that overwhelms the cooling system at 3C.

The practical implication: the maximum sustainable fast charge rate in a production EV is set not by the cell chemistry alone but by the combination of cell DCIR, cooling architecture, and BMS thermal strategy. The same cell in different packs can have very different practical fast charge limits.

What limits fast charge rate in production vehicles:

1. Cell maximum temperature — at high C-rates, temperature rises faster than the cooling system can reject heat
2. Cooling system capacity — cooling plate heat transfer coefficient × available temperature differential (coolant vs cell)
3. BMS thermal derate — the BMS reduces charge current as cell temperature approaches the limit, creating the characteristic tapering of charge rate visible on a charge curve
4. Pre-conditioning — without pre-warming cells to ≥15°C, peak charge rates cannot be achieved safely

→ Cell to pack fast charging thermal analysis → Tesla fast charge thermal model → Fast charging

Storage Temperature

Cells stored for extended periods outside the optimal temperature and SoC range will degrade faster than cells stored under optimal conditions.

Optimal storage conditions for lithium-ion:

• Temperature: 5°C to 20°C
• SoC: 30–50%
• Away from direct sunlight and heat sources

Why storage temperature matters: Calendar ageing continues in storage — the SEI layer grows slowly at the anode surface even without cycling. Higher storage temperature means faster SEI growth, higher self-discharge, and greater irreversible capacity loss. A cell stored at 40°C for one year will lose more capacity than the same cell stored at 20°C, even if neither is cycled.

For packs that may be stored for extended periods (seasonal BESS, service spares, manufacturing buffer stock), specifying and enforcing storage temperature is as important as specifying charge/discharge limits.

Related Sections

• Battery Cell — cell heat generation, specific heat capacity, and thermal conductivity are cell-level properties
• Battery Pack — cooling architecture, TIM selection, and enclosure thermal design are pack-level decisions
• BMS — the BMS controls the BTMS, monitors cell temperatures, and implements thermal derate
• Charging — fast charging is the most thermally demanding operational mode; thermal constraints set the practical charge rate limit
• Safety — thermal runaway and propagation are the thermal failure modes that pack design must contain
• Lifetime — operating temperature is the most controllable variable affecting cell ageing rate
• Mechanical — cell compression pressure affects thermal conductivity; cooling plate integration is a structural decision
• Manufacturing — TIM application, cooling plate assembly, and temperature management during formation are manufacturing considerations
• Modelling — thermal models of cells, modules, and packs are essential design tools
• Pack Sizing — thermal sizing determines whether passive or active cooling is needed and establishes cooling circuit requirements

→ Thermal Definitions & Glossary

See also: BYD Battery Self Heating System, Auxiliary Systems Power Consumption in Electric Vehicles, Heat Sources and Paths in a Battery Pack