The transition from internal combustion engine vehicles to electric vehicles (EVs) is accelerating globally, driven by the urgent need to address environmental pollution and energy security concerns. Among various new energy vehicle technologies, battery electric vehicles (BEVs) have gained significant traction due to their zero tailpipe emissions and relatively mature technological foundation. A critical factor influencing widespread consumer adoption is the driving range and the time required to replenish that range. Consequently, developing efficient high-power DC fast-charging solutions is paramount. However, high-power charging induces substantial heat generation within the lithium-ion battery pack. Excessive temperature or large temperature gradients can severely degrade battery performance, accelerate aging, and pose safety risks. Therefore, an effective Battery Thermal Management System (BMS), or more precisely, the thermal management subsystem coordinated by the BMS, is indispensable for enabling safe and efficient fast charging.
Existing research has primarily focused on optimizing charging protocols and improving onboard thermal management hardware. Charging strategies such as multi-stage constant current (MCC) and pulse charging have been developed to reduce polarization effects and control temperature rise. For thermal management, liquid cooling systems with optimized cold plate designs are prevalent for their high heat transfer efficiency. However, a significant limitation persists: during high-power charging, especially under extreme ambient temperatures, the onboard thermal management system must first condition the battery to an optimal temperature window (typically 20°C to 35°C) before applying peak charging currents. This preconditioning process, whether heating in cold climates or cooling in hot ones, consumes a considerable portion of the total charging time when using the vehicle’s own limited-capacity heat pump or chiller system.
This paper proposes a novel thermal management architecture designed to minimize this preconditioning overhead during high-power DC fast charging. The core concept involves transferring the primary thermal load from the vehicle to the charging infrastructure. We designed a charging-station-based heat pump system that pre-conditions a coolant reservoir to the desired temperature. When an EV connects for charging, this pre-conditioned coolant is directly circulated through the vehicle’s battery cooling circuit, enabling almost instantaneous thermal readiness. This system is referred to herein as the “Station-Based Thermal Management System” (S-BTMS), in contrast to the conventional “Onboard Thermal Management System” (O-BTMS).
The primary objective of this work is to quantitatively evaluate the benefits of the proposed S-BTMS in terms of charging time reduction and temperature control under various ambient conditions. Using numerical simulation, we model the electro-thermal behavior of a representative EV battery pack during a high-power charging process governed by a temperature-dependent charging map. The performance of the S-BTMS is compared against a baseline O-BTMS across five ambient temperatures: -20°C, -10°C, 0°C, 20°C, and 40°C. The analysis focuses on the battery’s temperature evolution, the applied C-rate, and the total charging duration from 0% to 100% State of Charge (SOC).
System Architecture Design
The efficacy of any fast-charging operation is heavily dependent on the integrated strategy between the charging protocol and the thermal management system. The Battery Management System (BMS) plays the central role in orchestrating this by monitoring cell voltages, temperatures, and SOC, then regulating both the current input from the charger and the operation of the thermal management components.
Conventional Onboard System (O-BTMS)
The baseline system, common in modern EVs, consists of two main loops: the battery coolant loop and the onboard heat pump loop. The battery loop contains a coolant (typically a glycol-water mixture) that circulates through cold plates attached to the battery modules. The heat pump loop, an air-source reversible system, provides heating or cooling capacity. A four-way reversing valve switches the heat pump between heating and cooling modes. Heat exchange between the two loops occurs via a liquid-to-refrigerant heat exchanger (chiller in cooling mode, condenser in heating mode). The BMS activates this system based on battery temperature, aiming to bring and maintain the pack within its optimal temperature window. However, the system’s capacity is limited by vehicle packaging, weight, and power consumption constraints, leading to relatively slow thermal response times, particularly during extreme temperature preconditioning.
Proposed Station-Based System (S-BTMS)
The proposed S-BTMS decouples the high-capacity thermal conditioning from the vehicle. Its architecture comprises three key circuits:
- Charging Station Heat Pump Loop: A relatively high-capacity, grid-powered heat pump system located within the charging station.
- Station Coolant Reservoir Loop: An insulated tank and pump system that holds a significant volume of coolant. This loop interfaces with the station’s heat pump to pre-heat or pre-cool the coolant to a setpoint temperature (e.g., 40°C for heating, 15°C for cooling) before a charging session begins.
- Vehicle Battery Coolant Loop: The vehicle’s existing cooling circuit with its cold plates and pump.
During a fast-charging session, a quick-connect coupling links the station’s reservoir loop directly to the vehicle’s battery loop. Valves within the vehicle isolate the onboard heat pump system. The pre-conditioned coolant from the station is then circulated through the battery cold plates at an optimal flow rate. This provides an immediate and powerful source of heat (in cold weather) or cooling (in hot weather), drastically reducing the time needed for the battery to reach its optimal charging temperature. The BMS’s role adapts to primarily manage the coolant flow and monitor temperatures, while the heavy lifting of thermal energy transfer is handled by the infrastructure. A comparative summary is presented in Table 1.
| Feature | Onboard System (O-BTMS) | Station-Based System (S-BTMS) |
|---|---|---|
| Primary Thermal Source | Vehicle heat pump/chiller (limited capacity) | Station heat pump/chiller (high capacity) |
| Pre-conditioning Energy Source | Vehicle battery (reduces range) | Grid power (no range impact) |
| Thermal Response Time | Slow, especially at extreme temperatures | Very fast, immediate coolant at target temperature |
| Vehicle Complexity & Weight | Higher (includes full heat pump system) | Potentially lower (smaller onboard system possible) |
| Infrastructure Requirement | Standard charging connector only | Charging connector + thermal coolant coupling |
| BMS Control Focus | Manage full thermal system, coordinate with charging | Manage battery loop, interface with station supply |
Battery Electro-Thermal Modeling
To accurately simulate the fast-charging process, a coupled electro-thermal model for the lithium-ion battery is essential. The model must account for the heat generation during charging and its effect on temperature, which in turn influences internal resistance and the allowable charging current dictated by the BMS.
Governing Thermal Equations
The temperature change within a battery cell is governed by the balance between heat generation, heat storage, and heat dissipation. The general energy balance for a battery can be expressed as:
$$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (\lambda \nabla T) + \dot{q}_{gen} $$
where $\rho$ is the average density ($kg/m^3$), $c_p$ is the average specific heat capacity ($J/kg·K$), $T$ is temperature (K), $t$ is time (s), $\lambda$ is the thermal conductivity tensor ($W/m·K$), and $\dot{q}_{gen}$ is the volumetric heat generation rate ($W/m^3$).
For a lumped-capacitance model applied to a single cell or module, and considering convective cooling from the cold plate, the equation simplifies to:
$$ m c_p \frac{dT}{dt} = Q_{gen} – h A_s (T – T_{coolant}) $$
where $m$ is the mass (kg), $h$ is the convective heat transfer coefficient ($W/m^2·K$), $A_s$ is the surface area for heat transfer ($m^2$), and $T_{coolant}$ is the coolant temperature (K). $Q_{gen}$ is the total heat generation rate in Watts.
Heat Generation Model
The heat generation within a lithium-ion battery during operation arises from irreversible (Joule) heating and reversible (entropic) heating. The widely used Bernardi model formulates the total heat generation rate as:
$$ Q_{gen} = I \left[ (V_{OCV} – V_t) – T \frac{\partial V_{OCV}}{\partial T} \right] $$
where $I$ is the current (A, positive for discharge, negative for charge), $V_{OCV}$ is the open-circuit voltage (V), $V_t$ is the terminal voltage (V), and $\frac{\partial V_{OCV}}{\partial T}$ is the entropy coefficient ($V/K$). The term $(V_{OCV} – V_t)$ represents the overpotential, which can be approximated as $I \cdot R_{int}$, where $R_{int}$ is the internal resistance. Thus, a more practical form for simulation is:
$$ Q_{gen} = I^2 R_{int}(SOC, T) + I T \frac{\partial V_{OCV}}{\partial T}(SOC) $$
The first term is always positive (Joule heating), while the second term can be positive or negative depending on the sign of the current and the entropy coefficient. During charging, if $\frac{\partial V_{OCV}}{\partial T} < 0$ (common for many chemistries), the entropic term absorbs heat, providing a cooling effect.
Determination of Battery Parameters
The model requires key input parameters. For a representative NCM (Nickel Cobalt Manganese) prismatic cell, the average properties are calculated using a mass-weighted average of constituent materials (cathode, anode, separator, electrolyte, casing).
Specific Heat Capacity ($c_p$):
$$ c_p = \frac{\sum m_i c_{p,i}}{\sum m_i} $$
Density ($\rho$):
$$ \rho = \frac{\sum m_i}{\sum V_i} $$
Thermal Conductivity: Due to the layered structure, conductivity is anisotropic. In-plane ($\lambda_x, \lambda_y$) and through-plane ($\lambda_z$) conductivities are calculated based on parallel and series thermal resistance networks, respectively:
$$ \lambda_{x,y} = \frac{\sum \lambda_i A_i}{\sum A_i}, \quad \lambda_z = \frac{\sum L_i}{\sum (L_i / \lambda_i)} $$
where $A_i$ is the cross-sectional area and $L_i$ is the thickness of layer $i$.
Critical electrochemical parameters like $R_{int}(SOC, T)$ and $V_{OCV}(SOC)$ are determined experimentally. Internal resistance is measured using a Hybrid Pulse Power Characterization (HPPC) test across different SOCs and temperatures. The $V_{OCV}$-SOC relationship is obtained by measuring the stabilized voltage after a long rest period at various SOC points. These relationships are fundamental inputs for the BMS’s state estimation and are crucial for an accurate simulation of the charging process.
Model Validation
The electro-thermal model was validated against experimental discharge data from literature for a similar cell. Under constant ambient temperature (30°C) and different discharge C-rates (0.5C, 1C, 2C, 3C), the simulated temperature rise profiles showed excellent agreement with the measured data, with a maximum error of less than 2.7% across all tests. This confirms the model’s adequacy for performance analysis and predictive studies of the battery’s thermal behavior under high-power charging.
Simulation Setup and Charging Strategy
We constructed a simulation model for a full vehicle battery pack and the two thermal management systems using a 1D system simulation software (AMESim).
Battery Pack Configuration
The pack is based on 96 prismatic NCM cells. Cells are arranged into 12 standard VDA590 modules (8 cells in series per module), with modules placed in two rows. Each module is cooled by an aluminum cold plate attached to its base. The cold plate features a simple serpentine flow channel. Key pack parameters are summarized in Table 2.
| Parameter | Value |
|---|---|
| Cell Chemistry | NCM/Graphite |
| Cell Capacity | 155 Ah |
| Nominal Cell Voltage | 3.7 V |
| Pack Configuration | 2P96S (355 V, ~55 kWh) |
| Avg. Density ($\rho$) | 2000 kg/m³ |
| Avg. Specific Heat ($c_p$) | 1138 J/kg·K |
| Thermal Conductivity ($\lambda_x, \lambda_y, \lambda_z$) | 19, 9.8, 5.9 W/m·K |
| Cooling Method | Liquid cooling (cold plate bottom) |
Thermal Management System Control
In both O-BTMS and S-BTMS simulations, the coolant pump is activated when the maximum battery temperature ($T_{max}$) exceeds a threshold (e.g., 25°C for cooling) or falls below one (e.g., 15°C for heating). The primary difference is the inlet coolant temperature ($T_{in}$):
- O-BTMS: $T_{in}$ starts at the ambient temperature and is gradually changed by the vehicle’s heat pump. Its rate of change is limited by the heat pump’s capacity.
- S-BTMS: $T_{in}$ is instantly set to a pre-defined optimal value (40°C for heating scenarios, 15°C for cooling scenarios) at the start of charging, simulating the connection to the station’s pre-conditioned reservoir.
Temperature-Dependent Charging Strategy
The BMS follows a pre-defined charging current map (C-rate vs. SOC and Temperature). This map is designed to maximize charging speed while respecting safety limits (cell voltage ≤ 4.2V, $T_{max}$ ≤ 45-50°C). It typically features:
- Low C-rate at very low temperatures (<10°C) to prevent lithium plating.
- Peak C-rate (e.g., 3.8C) within the optimal temperature window (20-35°C) and mid-SOC range.
- Tapered C-rate as SOC approaches 100% (constant-voltage phase) and/or if temperature approaches the upper safety limit.
The BMS dynamically selects the current from this map based on real-time $T_{max}$ and SOC. If $T_{max}$ exceeds a safety threshold (e.g., 45°C), the BMS can override the map to a lower C-rate and command maximum cooling.
Simulation Results and Analysis
Simulations were conducted for a full charge from 0% to 100% SOC under five ambient temperatures: -20°C, -10°C, 0°C, 20°C, and 40°C. The performance of the S-BTMS is compared directly against the O-BTMS.
Battery Temperature and C-rate Profiles
1. Low-Temperature Conditions (-20°C, -10°C, 0°C):
In all low-temperature cases, the O-BTMS requires a significant period to heat the battery from the sub-zero ambient to above 10-15°C before the BMS allows medium-to-high C-rate charging. The S-BTMS, by supplying 40°C coolant immediately, raises the battery temperature to the optimal zone within minutes. This allows the BMS to apply the peak 3.8C charge rate much sooner. As charging progresses and internal heat generation becomes significant, the systems eventually switch to cooling mode to prevent overheating. The S-BTMS consistently maintains a faster thermal response.
2. Mild-Temperature Condition (20°C):
At this near-optimal ambient temperature, both systems start with high C-rates immediately. The thermal trajectories are similar, as the primary heat source is internal generation. Both systems activate cooling midway to keep $T_{max}$ below 45°C. The advantage of the S-BTMS is minimal in this scenario, as little preconditioning is needed.
3. High-Temperature Condition (40°C):
The battery starts already warm. The O-BTMS struggles: the initial high C-rate causes $T_{max}$ to quickly hit the 45°C limit, forcing the BMS to drastically reduce the C-rate until the onboard cooling system can bring the temperature down. This results in a prolonged period of low-rate charging. The S-BTMS, by immediately supplying 15°C coolant, provides aggressive cooling from the start. This allows the BMS to sustain a higher average C-rate for longer and recover to a higher rate after a temperature-induced reduction, effectively controlling the peak temperature to a lower value (46°C vs 49.5°C for O-BTMS).
Charging Time Reduction Analysis
The total charging time and its breakdown by SOC phase are the most critical metrics. Table 3 summarizes the comparative results.
| Ambient Temp. | System | Charging Time per SOC Phase (min) | Total Time (min) | Time Reduction | ||
|---|---|---|---|---|---|---|
| 0-20% | 20-80% | 80-100% | ||||
| -20°C | O-BTMS | 28.4 | 11.8 | 14.2 | 54.4 | 24.6% |
| S-BTMS | 15.7 | 11.2 | 14.1 | 41.0 | ||
| -10°C | O-BTMS | 20.1 | 11.7 | 14.1 | 45.9 | 17.9% |
| S-BTMS | 12.4 | 11.2 | 14.1 | 37.7 | ||
| 0°C | O-BTMS | 12.3 | 11.4 | 14.1 | 37.8 | 10.6% |
| S-BTMS | 8.7 | 11.0 | 14.1 | 33.8 | ||
| 20°C | O-BTMS | 3.2 | 12.2 | 14.0 | 29.4 | 1.4% |
| S-BTMS | 3.2 | 11.8 | 14.0 | 29.0 | ||
| 40°C | O-BTMS | 4.0 | 17.1 | 14.0 | 35.1 | 6.0% |
| S-BTMS | 3.8 | 15.0 | 14.2 | 33.0 | ||
The results clearly demonstrate the significant advantage of the S-BTMS under temperature extremes:
- At -20°C: The S-BTMS reduces total charging time by 13.4 minutes (24.6%). The saving is almost entirely in the 0-20% SOC phase, where preconditioning dominates.
- At -10°C & 0°C: Savings of 17.9% and 10.6%, respectively. The benefit diminishes as the ambient temperature approaches the optimal range because the required heating energy decreases.
- At 20°C: Both systems perform nearly identically, with only a 0.4-minute (1.4%) difference, as minimal thermal intervention is needed.
- At 40°C: The S-BTMS saves 2.1 minutes (6.0%). The savings are concentrated in the 20-80% SOC phase, where its superior cooling capability allows the BMS to maintain a higher average C-rate compared to the O-BTMS, which is constrained by its slower cooling response.
This analysis underscores that the primary value of the infrastructure-based thermal management system is in drastically reducing the “thermal preparation” time at the beginning of a fast-charging session, particularly in cold climates. It also provides more effective temperature control in hot climates, enabling a marginally faster charge.
Conclusion
This study designed and analyzed a novel Station-Based Battery Thermal Management System (S-BTMS) for high-power DC fast charging of electric vehicles. The system shifts the primary thermal conditioning load from the vehicle’s limited-capacity system to a higher-capacity, grid-powered heat pump at the charging station. By delivering pre-conditioned coolant directly to the battery pack at the start of a charging session, it enables almost instantaneous thermal readiness.
Through detailed electro-thermal simulation across a range of ambient temperatures (-20°C to 40°C), the S-BTMS demonstrated substantial advantages over a conventional Onboard System (O-BTMS):
- Significant Charging Time Reduction in Cold Weather: The most dramatic improvements were seen at low temperatures. At -20°C, the S-BTMS reduced total charging time by 24.6%, with savings primarily in the critical 0-20% SOC phase where battery preconditioning typically occurs.
- Effective High-Temperature Management: At 40°C, the S-BTMS provided more aggressive cooling, controlling the peak battery temperature better (46°C vs. 49.5°C) and enabling a 6% faster charge by allowing the BMS to sustain higher charging currents for longer periods.
- Minimal Impact at Optimal Temperatures: At a mild 20°C, where little thermal management is needed initially, both systems performed similarly, confirming that the S-BTMS does not introduce inefficiencies under normal conditions.
The proposed architecture highlights a promising pathway for overcoming one of the key bottlenecks in EV fast charging: thermal constraints. By leveraging infrastructure-based thermal energy, the vehicle’s Battery Management System can execute an optimized high-power charging protocol with minimal delay, enhancing the consumer experience, especially in regions with extreme climates. Future work could focus on the detailed design of the coupling interface, optimization of the station-side energy management, and a comprehensive cost-benefit analysis of deploying such infrastructure. The integration of such a thermal management system represents a significant step towards making fast charging as convenient and time-efficient as refueling a conventional vehicle.