EV Battery Thermal Management: Silicone Heating Pad Integration for Lithium-Ion Pack Preconditioning at −30°C

A lithium-ion cell that sits at −30°C for an hour is, electrically speaking, a different component from the same cell at +25°C. Internal resistance climbs roughly 4–6× across that range, the electrolyte viscosity pushes the lithium-ion diffusion coefficient down by an order of magnitude, and lithium plating begins at the anode surface as soon as the cell accepts charge at any meaningful C-rate below 0°C. The cell does not recover from plating damage; it is permanent capacity loss that accumulates cycle by cycle. Because the cell cannot accept charge safely below 0°C, so the EV thermal management system has to warm the pack from −30°C to at least +5°C to +10°C before the BMS will allow charging current. The component that does that warming is, in most modern EV architectures, a silicone rubber heater for battery integrated into the module.

This article is written for the EV thermal engineer, the BMS firmware engineer, and the pack integration supplier who has to specify, qualify, and validate a silicone heating pad for a production lithium-ion pack. It draws on the design and qualification work our team at JINGWEI has supported across passenger-EV, hybrid, and commercial-vehicle programs — from the pad’s etched-foil construction through the BMS PID loop to the UN 38.3 and ISO 12405-4 component tests. The numbers below are working values, not generic marketing ranges; verify them against your own pack thermal model and your own safety case before signing off on a production design.

Why a Cold Pack Needs a Heating Pad at All

The standard solution to cold-pack preconditioning in the global EV market is a heated coolant loop — a PTC heater in the coolant path warms the coolant, the coolant circulates through channels in the pack baseplate, and the cells warm from the bottom up. That architecture is well understood and dominates the long-range passenger-EV market. The battery heating silicone pad is the alternative architecture, dominant where one or more of the following constraints apply:

• No liquid cooling loop is available. Air-cooled packs, two-wheeler and three-wheeler battery packs, e-bike packs, light-EV packs, and most 48V mild-hybrid packs do not have a coolant circuit. A pad laminated directly to the cell or module baseplate is the only practical heat source.
• The pack must precondition without a thermal-management pump running. Some EV architectures require preheat before the coolant pump is energised, either to reduce start-up load on the 12V bus or to precondition the pack while the vehicle is plugged in and the HV contactors are open. A pad that runs directly from the HV bus (with a dedicated DC-DC converter) does not need the pump to be awake.
• The pack has a strict Z-height budget. A silicone pad is 1.5–3.0 mm thick; a coolant channel with a baseplate is 8–15 mm. For a pack where the module height is constrained by the vehicle floor, the pad saves 5–12 mm of Z-height that the cell stack can use.
• The pack uses pouch or prismatic cells with large face areas. Cell-face heating is a natural fit for the pad architecture; coolant-baseplate heating leaves the cell face under-utilised as a heat-transfer surface.

Because the pad is in direct contact with the cell or the baseplate, so the heat-transfer path is short, the response time is fast, and the temperature uniformity across the cell face is much better than a baseplate-only architecture. The trade-off is that the pad has a smaller thermal mass than a coolant loop and is correspondingly less effective at absorbing short-duration high-power heat loads. For preconditioning, where the goal is to warm a cold pack at low power over 30–45 minutes, the pad is the right answer.

Inside a Silicone Heating Pad: Construction and Materials

A production-grade EV battery silicone heating pad is a five-layer laminate, typically 1.5–3.0 mm thick. From the cell-facing side outward:

1. Adhesive / thermal-interface layer (50–150 µm). A pressure-sensitive acrylic adhesive (PSA) with a thermal conductivity of 0.8–1.6 W/m·K, or a thin silicone bond layer for higher-temperature programs. The PSA has to bond at room temperature and survive thermal cycling between −40°C and +85°C without creeping or losing shear strength.
2. Inner silicone rubber sheet (0.5–1.0 mm). Fibreglass-reinforced silicone rubber, providing dielectric strength and a thermal-spread layer. The fibreglass scrim limits the coefficient of thermal expansion and prevents the silicone from cold-flowing under the bond.
3. Heating element (50–200 µm). Either an etched-foil resistance circuit (most common for EV) or a wire-wound element. Etched-foil gives a uniform watt density and a tight resistance tolerance (typically ±5%); wire-wound is cheaper but gives a stripe pattern that needs to be considered in the cell-face temperature map.
4. Outer silicone rubber sheet (0.5–1.0 mm). Same construction as the inner sheet, with a flame-retardant grade (UL 94 V-0) for the surface facing the module housing or the foam insulation.
5. Insulation jacket (optional, 0.5–1.0 mm). Closed-cell silicone foam or aerogel blanket on the outer face, to reduce heat loss into the air gap and direct the heat flux toward the cell. Skipped on baseplate pads where the heat flow is already directed downward.

Because the etched-foil pattern is the only active heat-generating layer, so the design decisions that matter most are the foil thickness (typically 0.05–0.10 mm of nickel-chromium alloy or stainless steel 304), the circuit track width and pitch (which together set the watt density), and the termination method (which sets the dielectric withstand). Get any of those wrong and the pad will either under-perform, over-heat locally, or fail the dielectric-withstand test. The supply-chain risk on a production EV program is that the foil alloy and the silicone rubber grade are both single-sourced; a second-source qualification is part of the component release, not a nice-to-have.

Watt Density and Power Sizing for a 60 kWh Pack

Watt density is the single design parameter that the thermal engineer has to get right before the pad is laid out. The density is watts per square centimetre of active pad area, and it is set by three constraints: the cell face area available for the pad, the maximum cell-face temperature allowed by the cell supplier, and the time budget for the cold-start.

Because the cell face is the most temperature-sensitive surface, so a cell-face pad is held below 0.6 W/cm² to keep the local cell-face temperature rise inside the 5°C delta the BMS expects during preconditioning. A baseplate pad can run higher because the baseplate spreads the heat before it reaches the cell. Exceed 1.5 W/cm² on a cell face and the local hot-spot will trigger the BMS protection circuit and abort the preconditioning cycle, which is the opposite of what the driver wants in −30°C weather.

Integration: Pad-to-Module, Pad-to-BMS, Pad-to-Pack

Three integration decisions determine whether a correctly designed pad works in the field.

1. Mechanical attachment to the module. The PSA bond works for hand-built prototypes and small-series packs, but for volume production a mechanical clamp is preferred. A spring-clip or compression-frame design that maintains 0.2–0.5 MPa of contact pressure across the pad area survives thermal cycling better than a PSA-only design, and it can be reworked in the field if a cell is replaced. A peel strength of 5–8 N/cm on the PSA at room temperature is the minimum; below that, the pad will creep over time.
2. Electrical termination. The lead wires exit the pad in a strain-relieved overmould, terminate to a connector on the module housing, and run back to a pad-controller PCB. The connector must be a sealed automotive-grade type (USCAR, LV214, or equivalent) and must be keyed to prevent reverse polarity. The pad-controller PCB is the interface between the BMS and the pad; it carries the solid-state relay, the current sensor, and the temperature feedback.
3. Thermal interface to the cell or baseplate. The PSA’s thermal conductivity is the dominant heat-transfer bottleneck. A standard acrylic PSA is 0.8–1.2 W/m·K; a filled PSA with ceramic or graphite is 1.5–2.5 W/m·K. The cost difference is small relative to the pack-level gain, and a 1.0 W/m·K upgrade can shave 8–12 minutes off the cold-start time at the same watt density. Because the cold-start time is the user-visible metric that drives customer satisfaction, so the PSA choice belongs in the design review, not the procurement review.

One more integration point that catches first-time pack designers: the pad’s lead wires are inside the high-voltage enclosure. They must be rated to the same dielectric withstand as the HV bus (typically 2,500 V AC for 1 minute between conductor and the module housing) and the cable jacket must be a low-smoke zero-halogen compound to meet the pack-level fire and smoke requirements. Standard UL 2464 hookup wire is not acceptable for this application.

BMS Control Logic: The PID Loop That Protects the Cell

The pad is dumb. The BMS is smart. The control loop has three responsibilities, in priority order:

1. Never allow the cell face to exceed the supplier’s published maximum temperature. A typical NMC cell has a maximum cell-face temperature of 60°C during normal operation and 80°C during fault conditions. The pad must shut off well before either limit; a working setpoint is 45°C for normal preconditioning and 55°C as the hard upper limit.
2. Maintain the cell in the +5°C to +10°C window during the entire charge cycle. Below +5°C, the BMS should keep the pad on. Above +10°C, the BMS should switch the pad off and let the cell self-warm. A typical PID loop runs at 1–2 Hz with a hysteresis of 2–3°C to avoid relay chattering.
3. Fault-detect the pad on every cold-start cycle. Open-circuit, short-circuit, and over-current faults must be detected within 100 ms and the pad disabled. The pad-controller PCB carries the current sensor for this; the BMS reads the fault flag and the pack controller decides whether to abort the cold-start or continue with reduced power.

Because the pad is inside the HV enclosure, so the BMS firmware has to assume the pad could fail in a closed (always-on) state. The mitigation is a hardware safety cut-off: a normally-open contactor in series with the pad, controlled by the BMS safety line. The contactor is held closed during normal operation and is the single point that the BMS can de-energise in a thermal-runaway event. The contactor is sized to break the full pad current at the worst-case locked-rotor condition, and the contactor status is read back to the BMS on a dedicated GPIO.

Test & Validation: UN 38.3, ISO 12405-4, IEC 62660-2

Three test documents govern a production pad. The OEM will ask for all three; the pad supplier should be able to produce all three without re-testing.

• UN 38.3 (transport). The UN Manual of Tests and Criteria, Part III, Section 38.3. The pad itself is not a cell, but it is treated as a component of the pack and is subject to the altitude simulation, thermal cycling, vibration, shock, external short-circuit, impact, overcharge, and forced-discharge tests as they apply to a resistive heating element. A typical pass is T1–T8 with a +55°C to −40°C thermal cycle and a 13 kN shock.
• ISO 12405-4 (performance). The international standard for lithium-ion traction battery packs. ISO 12405-4 covers the pack-level performance tests, and the pad is qualified as a component during the pack-level cold-start test, the preconditioning test, and the energy-efficiency test. The pad supplier provides a component-level test report that the pack integrator rolls up into the pack-level submission.
• IEC 62660-2 (reliability). The reliability and abuse testing standard for lithium-ion cells for EV propulsion. IEC 62660-2 covers the cell-level tests, and the pad is qualified by demonstrating that it does not cause a cell to fail any of the abuse tests. A pad that hot-spots a cell during the thermal-runaway test is a pad that has failed the program.

Because the three standards overlap but are not identical, so the component-level test plan for a production pad has to cover the union of all three. The cheapest way to do this is to run a single DV/PV test sequence that satisfies the strictest requirement for each test item, then have the test report cover all three. The supplier that has already done this for an OEM program can hand over the report in days; the supplier that has to re-test from scratch is a six-month delay.

What Goes Wrong in the Field, and How to Prevent It

Three failure modes account for the majority of field returns on EV battery heating pads:

• PSA creep at high temperature. The pad is bonded to the cell face; the cell warms to 40°C during fast charging; the PSA softens; the pad creeps and exposes an air gap; the cell face cools non-uniformly. Prevention: use a silicone bond instead of an acrylic PSA for packs that fast-charge at high C-rates, or add a mechanical clamp to back up the PSA.
• Dielectric breakdown at the lead-wire exit. The lead-wire exit is the most highly-stressed point in the pad. A manufacturing defect in the overmould allows the silicone to thin, the dielectric withstand fails the 2,500 V AC test, and the BMS reports an isolation fault. Prevention: 100% dielectric-withstand test on every pad at the supplier’s outgoing inspection, and a documented overmould process capability index (Cpk ≥ 1.33).
• Watt-density drift over the pad lifetime. The etched-foil element ages; the resistance climbs by 5–10% over 10 years; the watt density drops proportionally; the cold-start time lengthens. Prevention: design the pad with a 20% watt-density headroom on day one, and specify a maximum resistance drift in the supplier quality agreement.

A correctly designed pad is invisible to the driver. It runs during a cold-start, switches off when the cell reaches +10°C, and is silent for the next 7–10 years. The pads that fail in the field are the ones that were over-specified on day one (oversized watt density, undersized dielectric margin) or under-specified on the integration (PSA-only bond, no overmould, no fault detection). The cost of a field failure on a single pack is roughly 100x the cost of getting the pad right at the design stage, which is why the seven sections above matter more than the unit price.

Frequently Asked Questions

Where to Go From Here

If you are sizing a pad for a new pack, the next step is to share three numbers with the pad supplier: the available cell-face or baseplate area per module, the maximum cell-face temperature allowed by the cell supplier, and the cold-start time budget. The silicone rubber heater for battery product line carries the watt-density range and the dielectric test report referenced in Sections 3 and 4. The battery heating silicone mat product page has the standard sizes and the lead-wire options. For a custom watt-density or a custom outline, the JINGWEI EV heater solutions team can quote a DV sample in 10–15 working days from a confirmed spec sheet.