Model-based upscaling of vanadium redox flow battery systems: engineering challenges and solutions

1. Problem Statement

The integration of stochastic renewable energy sources into the electric grid requires large-scale energy storage. While Vanadium Redox Flow Batteries (VRFBs) offer advantages such as long lifetimes and independent scalability of power and capacity, their commercial viability at the grid scale is hindered by efficiency losses and thermal management challenges.

• Thermal Constraints: VRFB electrolytes must be maintained between 5–40 °C. Temperatures exceeding 40 °C cause vanadium ion precipitation, while low temperatures reduce kinetics.
• Scaling Complexity: As systems scale from small (kW) to large (hundreds of kW), the interaction between electrical, chemical, hydraulic, and thermal processes becomes non-linear. Existing models often lack the granularity to predict how system efficiency changes with different configurations (cell numbers, stack arrangements, and current densities) in real-world, containerized environments.
• Efficiency Gap: While electrochemical efficiency is high, the net system efficiency is significantly reduced by auxiliary loads (pumps, inverters, and thermal management systems), which vary depending on the specific upscaling configuration.

2. Methodology

The authors developed a comprehensive multi-physics lumped parameter model in Matlab/Simulink to simulate a containerized VRFB system.

• Model Architecture:

  • Thermal: A detailed thermal network with 14–27 nodes (depending on stack count) modeling the stack, tanks, pipes, inner air, container walls, and auxiliary equipment. It accounts for convection, conduction, and a dynamic radiative heat transfer network using view factors calculated via a 3D Ansys Fluent model.
  • Electrical/Chemical: Uses State-of-Charge (SoC) dependent resistance and Nernst equations. It models entropic heat (reversible) and Joule heating (irreversible), as well as side reactions causing self-discharge.
  • Hydraulic: Calculates pump head and power consumption based on pressure losses in manifolds, cells (interdigitated flow field), and pipes.
  • Thermal Management System (TMS): A hybrid active system including openable dampers, cooling fans, an Air Conditioner (AC), and periodic electrolyte flushing (standby mode) to manage heat.

• Simulation Scope:

  • Scaling Cases: 10 distinct power cases ranging from 4 kW to 400 kW (following the Renard series), with a fixed nominal capacity-to-power ratio of 4 (e.g., 4 kW/16 kWh to 400 kW/1600 kWh).
  • Configurations: For each power case, 18 different configurations were tested by varying the number of cells and the discharging current (limiting max current to 50–250 A). This resulted in 180 total configurations.
  • Environmental Data: Simulations used real weather data (Budapest) with 1-minute resolution for ambient temperature and solar irradiance over a 20-day period.

3. Key Contributions

• Detailed Radiative Thermal Model: Unlike previous studies, this model explicitly calculates radiative heat exchange between all internal components (stacks, tanks, walls) and the container roof, which is critical for large-scale systems with high surface areas.
• Hybrid TMS Strategy: The study evaluates a sophisticated TMS logic that switches between fans, AC, and electrolyte flushing based on temperature thresholds and Coefficient of Performance (COP) optimization.
• Comprehensive Upscaling Analysis: It provides a unique dataset linking system efficiency directly to the ratio of discharging current vs. cell number, revealing how different architectural choices impact net efficiency.
• Containerized Reality: The model incorporates the thermal inertia of the container, solar radiation absorption on walls, and the specific layout of pumps and stacks within a shipping container.

4. Key Results

A. Efficiency Trends

• Electric Efficiency: Ranges from 68% to 89%. It is primarily driven by ohmic losses; lower discharging currents yield higher electric efficiency. The number of cells (output power) has a negligible effect on electric efficiency at a fixed current.
• System Efficiency: Ranges from 43% to 66%. This is significantly lower than electric efficiency due to auxiliary loads.
  • Low-current configurations generally achieve higher system efficiency.
  • High-power output configurations (more cells) show improved system efficiency compared to low-power configurations at the same current, due to better distribution of auxiliary loads.

B. Loss Breakdown

1. Thermal Management System (TMS): The largest source of loss, ranging from 7.5% to 17% in low-power systems and 6% to 14.2% in high-power systems.
   • The AC is the dominant consumer, especially in high-current or high-temperature scenarios.
   • Medium-current configurations (e.g., 70–110 A) often achieve the optimal balance, minimizing TMS losses by balancing Joule heat, entropic heat, and self-discharge heat.
2. Inverter Losses: Consistent reduction of 11–12% across configurations.
3. Pump Losses:
   • Single-stack: Losses are higher at low power (up to 6.5%) because pump minimum power consumption is a larger fraction of total output.
   • Multi-stack: Losses drop to 1–2% as the system scales up, as pumps are distributed (one per stack) and operate closer to optimal points.

C. Thermal Dynamics

• Self-Discharge: During standby, exothermic self-discharge reactions generate heat. Without periodic flushing, this can raise stack temperatures significantly.
• Radiation: In large multi-stack systems, the top stack (closest to the roof) becomes the hottest due to solar absorption, creating temperature gradients of up to 5.7 °C between stacks in large configurations.
• Cooling Strategy: Periodic electrolyte flushing (circulation only during standby) reduces pump energy consumption by ~33% compared to continuous circulation while maintaining safe temperatures.

5. Significance and Implications

• Design Guidance: The study provides critical design rules for upscaling VRFBs. It suggests that low-current, high-cell-number configurations generally yield the highest net system efficiency, provided the thermal management is optimized.
• Thermal Management Optimization: The results highlight that a "one-size-fits-all" TMS is inefficient. The study advocates for dynamic flow factor modulation and two-stage pumping systems to exploit high COP during specific thermal conditions.
• Scalability Validation: The research confirms that while scaling up reduces relative pump losses, it increases the complexity of thermal uniformity. Proper spatial arrangement (e.g., placing stacks on the ground rather than near the roof) is essential to mitigate solar-induced temperature gradients.
• Future Directions: The authors propose that future work should focus on waste heat recovery (thermally regenerative cycles) and advanced control strategies for dynamic flow rates to further push system efficiency beyond the current 66% ceiling.

In conclusion, this paper bridges the gap between theoretical electrochemical models and practical engineering constraints, offering a validated framework for designing efficient, large-scale, containerized VRFB systems.