Thermal management has become a critical challenge in the large-scale deployment of battery energy storage systems. The accumulation of heat within battery packs, if not effectively dissipated, accelerates aging, shortens service life, and in severe cases triggers thermal runaway, leading to safety incidents. To simultaneously enhance the cooling effectiveness and reduce energy consumption, this paper proposes a novel indirect liquid-cooling system based on mechanical vapor recompression (MVR) and falling film evaporation. The system is designed specifically for battery energy storage systems, aiming to achieve high efficiency and low energy consumption. Using pure water as the working fluid, the MVR falling film evaporator operates under vacuum conditions to exploit the high heat transfer coefficient of phase-change evaporation. The compressor recovers and upgrades the latent heat of vapor, thereby eliminating the need for a conventional refrigeration cycle. This article presents the detailed flow scheme, simulates the thermodynamic cycle with Aspen Plus, compares its performance with a conventional air-conditioning refrigeration system under a cooling load of 210 kW, and analyzes the influence of key parameters. The results demonstrate that the proposed system can save up to 77.04% of the electricity consumption compared to a first-level energy-efficiency refrigeration system, and the investment payback period is only 2 months, highlighting its significant economic and energy-saving potential for battery energy storage systems.
Conventional cooling approaches for battery energy storage systems are mainly air cooling and liquid cooling. Air cooling is simple and low-cost, but its heat dissipation efficiency and temperature uniformity are inferior to those of liquid cooling. Liquid cooling, especially indirect cold-plate cooling, has been widely adopted in electric vehicles and small-scale storage stations. However, it requires complex flow channel design, continuous coolant filling, and an additional chiller unit, which increases system complexity and cost. Direct immersion cooling, although providing excellent temperature uniformity, uses expensive electronic fluorinated liquids, suffers from high coolant consumption, and demands rigorous pressure control. In all these schemes, the cooling capacity is typically supplied by an air-conditioning refrigeration unit that consumes substantial electrical power. The energy efficiency of the cooling system is often overlooked in existing studies, which mainly focus on heat dissipation capacity. Therefore, designing a more energy-efficient and cost-effective cooling architecture is of great importance for the sustainable deployment of battery energy storage systems.
This work introduces a closed-loop cooling system that integrates falling film evaporation with MVR technology. The core concept is to use water—a low-cost, high-specific-heat coolant—as the evaporating medium. The evaporation occurs on the outer surface of falling-film plates that are attached to the battery modules. The vapor is then compressed by an MVR compressor to increase its saturation temperature, enabling heat rejection to the ambient via a condenser. The condensed water is recirculated through a pressure-reducing valve to the evaporator. This approach eliminates the need for a separate refrigeration cycle, thus significantly reducing the compressor work compared to conventional vapor-compression refrigeration. The system operates under vacuum in the evaporator side to lower the boiling point of water, making it suitable for moderate cooling temperatures (e.g., 303 K). The following sections detail the system configuration, simulation methodology, performance comparison, and economic evaluation.
System Configuration and Working Principle
The proposed MVR falling film indirect liquid-cooling system consists of two main modules: the falling-film evaporator module and the MVR module. The evaporator module includes battery cells, liquid distributors, and falling-film plates that form a vacuum chamber. The MVR module comprises a compressor, a gas-liquid separator, a condenser, a pressure-reducing valve, circulating pumps, and control valves. The flowchart of the system is shown schematically in Figure 1.
The operation sequence is as follows: First, the entire system is evacuated, and then a predetermined amount of deionized water is injected. After reaching the desired vacuum level, the compressor and circulation pump are started. Water from the gas-liquid separator is pumped to the liquid distributor, which uniformly spreads a thin film on the falling-film plates adjacent to the battery modules. The water film absorbs the heat generated by the batteries and evaporates under vacuum conditions. The vapor flows to the gas-liquid separator due to the pressure gradient; any entrained liquid droplets are separated and returned to the distributor. The dry vapor then enters the compressor, where it is compressed to a higher pressure and temperature. The superheated vapor is condensed in the condenser by rejecting heat to an external cooling medium (e.g., ambient air or cooling water). The condensed liquid water passes through a pressure-reducing valve to drop its pressure to the saturation pressure corresponding to the evaporation temperature, and then it is recycled to the falling-film evaporator. The indirect contact design ensures that the batteries operate at atmospheric pressure, while the cooling medium is confined within the vacuum chamber, enhancing safety and reliability.
The falling-film plate structure is depicted in Figure 2. Each battery cell is attached to a conductive metal plate (stainless steel or aluminum) with thermal grease to minimize contact resistance. The plates are sealed to form a closed vacuum cavity. The liquid distributor at the top ensures even wetting of the plate surfaces, promoting stable film evaporation. This configuration eliminates the need for internal flow channels within the cold plate, reduces the thermal resistance and coolant inventory, and allows operation under constant ambient pressure for the battery modules.
Thermodynamic Simulation and Model Setup
The system was simulated using Aspen Plus with the IAPWS-95 property method for water and steam. The following assumptions were made to simplify the model: steady-state conditions, negligible pressure drops and heat losses in pipes, and no non-condensable gases. The evaporator was modeled as a black-box where all liquid water is assumed to evaporate completely. The compressor isentropic efficiency was set to 0.85 based on typical values for small-scale MVR applications.
The compressor power consumption is given by:
$$W = \frac{m_v \cdot \Delta H_{id,v}}{\eta_c}$$
where \(W\) (kW) is the compressor power, \(m_v\) (kg/s) is the mass flow rate of vapor, \(\Delta H_{id,v}\) (kJ/kg) is the isentropic enthalpy rise of the vapor across the compressor, and \(\eta_c\) is the isentropic efficiency. The evaporator cooling load is:
$$Q = m_w \cdot \Delta H_{w,e}$$
where \(Q\) (kW) is the cooling load, \(m_w\) (kg/s) is the water flow rate (equal to \(m_v\) under full evaporation), and \(\Delta H_{w,e}\) (kJ/kg) is the enthalpy change of water from liquid to vapor at the evaporation temperature.
The baseline design parameters are summarized in Table 1.
| Parameter | Value |
|---|---|
| Cooling load (target) | 210 kW |
| Cooling temperature (evaporation) | 303 K |
| Operating pressure (evaporator) | 4,247 Pa |
| Mass flow rate of water | 311 kg/h |
| Compressor temperature rise (ΔTcomp) | 15 K |
| Compressor isentropic efficiency | 0.85 |
The simulation results at the baseline conditions are shown in Table 2. The actual net cooling load from the evaporator was 204.492 kW (slightly lower than 210 kW due to assumed heat losses), and the compressor power consumption was only 14.227 kW.
| Parameter | Value |
|---|---|
| Net cooling load (Q) | 204.492 kW |
| Compressor power (W) | 14.227 kW |
| Latent heat of vaporization at 303 K | 2,429.67 kJ/kg |
| COP equivalent (Q/W) | 14.37 |
Comparative Analysis with Conventional System
A conventional air-conditioning refrigeration system designed for the same cooling load (210 kW) was selected as a reference. The reference system is a room-level chilled water unit with a coefficient of performance (COP) of 3.3 (first-level energy efficiency). Its power consumption is calculated as:
$$P_{conv} = \frac{Q}{COP} = \frac{210}{3.3} \approx 63.636 \text{ kW}$$
However, to match the actual net cooling load of 204.492 kW provided by the MVR system, the conventional system would consume 204.492/3.3 = 61.967 kW. The energy saving ratio is:
$$\text{Saving} = \frac{P_{conv} – W}{P_{conv}} \times 100\% = \frac{61.967 – 14.227}{61.967} \times 100\% = 77.04\%$$
Table 3 summarizes the comparison under identical effective cooling load.
| Item | MVR system | Conventional (COP=3.3) |
|---|---|---|
| Compressor/fan power (kW) | 14.227 | 61.967 |
| Annual electricity consumption (3,000 h/yr) | 42,681 kWh | 185,901 kWh |
| Energy saving ratio | — | 77.04% |
The MVR system achieves a 77% reduction in electricity consumption, primarily because it avoids the vapor-compression refrigeration cycle. In a conventional system, the compressor must handle the temperature lift from the evaporator (e.g., 7°C) to the condenser (e.g., 45°C), whereas in the MVR system, the compressor only needs to overcome the temperature difference between the evaporator (30°C) and the heat rejection temperature (e.g., 45°C) plus the approach temperature. Since the evaporation temperature in the MVR system is relatively high (30°C vs. 7°C for chilled water), the pressure ratio is much lower, leading to significantly less work.
Effect of Key Thermodynamic Parameters
Two critical parameters affecting system performance are the cooling temperature (evaporation temperature) and the compressor temperature rise. The cooling temperature determines the saturation pressure and vapor density, while the compressor temperature rise dictates the pressure ratio. Figures 3 and 4 in the original study show these influences; here we present the data in tabular and equation forms.
Table 4 lists the compressor power at different cooling temperatures and compressor temperature rises, as extracted from the simulation.
| Cooling Temp (K) | ΔT=10 K | ΔT=12 K | ΔT=15 K |
|---|---|---|---|
| 283 | 10.523 | 12.368 | 15.612 |
| 293 | 10.114 | 11.903 | 15.012 |
| 303 | 9.702 | 11.434 | 14.227 |
| 313 | 9.286 | 10.961 | 13.713 |
The compressor power decreases with increasing cooling temperature because the suction temperature is higher, reducing the pressure ratio and thus the specific work. This relationship can be approximated by a linear regression:
$$\frac{\partial W}{\partial T_{evap}} \approx -0.12 \text{ kW/K} \quad (\text{for } \Delta T = 15\text{ K})$$
Conversely, the compressor power increases with compressor temperature rise. For a cooling temperature of 303 K, the power varies as:
$$W = 0.904 \cdot \Delta T_{comp} – 0.317 \quad (\text{in kW, with } \Delta T_{comp} \text{ in K})$$
The energy saving ratio relative to the conventional system (COP=3.3) also varies with these parameters, as shown in Table 5.
| Cooling Temp (K) | ΔT=10 K | ΔT=12 K | ΔT=15 K |
|---|---|---|---|
| 283 | 82.2 | 80.0 | 74.8 |
| 293 | 83.0 | 80.8 | 75.8 |
| 303 | 83.9 | 81.5 | 77.0 |
| 313 | 84.8 | 82.3 | 78.1 |
The saving ratio increases with cooling temperature and decreases with compressor temperature rise. For instance, at a fixed ΔT of 15 K, raising the cooling temperature from 283 K to 313 K increases the saving ratio from 74.8% to 78.1%, an improvement of 3.3 percentage points. Reducing ΔT from 15 K to 10 K at 303 K increases the saving ratio from 77.0% to 83.9%, an improvement of 6.9 percentage points. Therefore, from an energy perspective, it is desirable to operate at the highest possible cooling temperature (within battery thermal limits) and the lowest feasible compressor temperature rise (limited by heat sink temperature).
The volumetric flow rate entering the compressor is another important factor. The suction volume flow rate is inversely proportional to the vapor density, which increases with cooling temperature. Table 6 gives the suction volume flow rate for different cooling temperatures at a fixed mass flow of 311 kg/h.
| Cooling Temp (K) | Vapor density (kg/m³) | Volume flow (m³/h) | Volume flow (m³/s) |
|---|---|---|---|
| 283 | 0.00958 | 32,470 | 9.02 |
| 293 | 0.01357 | 22,920 | 6.37 |
| 303 | 0.01878 | 16,560 | 4.60 |
| 313 | 0.02543 | 12,230 | 3.40 |
Higher cooling temperature reduces the volumetric flow rate, which not only lowers compressor power but also allows a smaller compressor size, reducing capital cost. However, the cooling temperature cannot be arbitrarily increased because battery energy storage systems typically have an optimal operating temperature range (e.g., 20–40°C). For many lithium-ion batteries, the maximum allowable temperature is around 45–50°C, so a cooling temperature of 303–313 K (30–40°C) is acceptable.
Economic Analysis
The economic viability of the proposed MVR system was evaluated using annualized cost (AC) and payback period (PBP). The annualized cost includes capital recovery, maintenance, and electricity consumption. The initial investment cost for the MVR system is given in Table 7, based on major components only (excluding piping, instrumentation, and installation).
| Component | Quantity | Unit Price (RMB) |
|---|---|---|
| Vapor compressor (screw type, 15 kW) | 1 | 12,000 |
| Condenser (shell-and-tube, 210 kW) | 1 | 3,000 |
| Gas-liquid separator | 1 | 1,000 |
| Pressure reducing valve | 1 | 400 |
| Circulation pump | 1 | 500 |
| Control valves, fittings | 1 | 300 |
| Total | 17,200 |
The conventional system (room-level chilled water unit with 210 kW cooling capacity) typically costs about 24,200 RMB (including compressor, evaporator, condenser, expansion valve, fans, etc.).
The annualized cost model used the following equations:
Annualized capital cost:
$$C_{acc} = C_{ccm} \times \frac{i(1+i)^n}{(1+i)^n – 1}$$
where \(C_{ccm}\) is the initial investment, \(i\) is the interest rate (10%), and \(n\) is the lifetime (15 years). Annual salvage value:
$$S_a = S \times \frac{i}{(1+i)^n – 1}$$
where \(S\) is the salvage value (20% of initial cost). Maintenance cost is taken as 10% of \(C_{acc}\). Annual electricity cost:
$$C_{rec} = C_{ee} \times W \times t$$
where \(C_{ee}\) is the electricity price (0.8 RMB/kWh), \(W\) is the system power consumption (kW), and \(t\) is the annual operating hours (3,000 h). The total annualized cost is:
$$AC = C_{acc} + C_{mc} – S_a + C_{rec}$$
The payback period is calculated using the discounted cash flow method:
$$PBP = \frac{\ln\left[1 – \frac{C_{ccm}}{B_j}(i – r)\right]}{\ln\left(\frac{1+r}{1+i}\right)}$$
where \(B_j\) is the first-year saving (difference in electricity cost between conventional and MVR systems), and \(r\) is the discount rate (6%).
Table 8 presents the economic comparison between the MVR system and the conventional refrigeration system.
| Item | MVR System | Conventional System |
|---|---|---|
| Initial investment (Cccm) | 17,200 RMB | 24,200 RMB |
| Annualized capital cost (Cacc) | 2,262 RMB | 3,182 RMB |
| Annual maintenance (Cmc) | 227 RMB | 319 RMB |
| Annual salvage value (Sa) | 86 RMB | 121 RMB |
| Annual electricity consumption (kWh) | 42,681 | 185,901 |
| Annual electricity cost (Crec) | 34,145 RMB | 148,721 RMB |
| Total annualized cost (AC) | 36,548 RMB | 152,101 RMB |
| First-year saving (Bj) | 114,576 RMB | |
| Payback period (PBP) | 0.163 years (approx. 2 months) | |
The MVR system has a lower initial cost because it eliminates the refrigeration cycle components (evaporator, expansion valve, and refrigeration-side piping). More importantly, the annual electricity saving is about 114,576 RMB, leading to a payback period of only 2 months. Even considering a 10% uncertainty in electricity price or component cost, the payback remains well under 1 year, making the MVR solution highly attractive for battery energy storage systems. Additional benefits include reduced maintenance complexity (no refrigerant handling) and improved safety due to operation at vacuum with water instead of flammable refrigerants.
Conclusions
This study presents a novel indirect liquid-cooling system for battery energy storage systems that combines falling film evaporation with mechanical vapor recompression. The system uses water as the working fluid under vacuum, eliminating the need for a conventional refrigeration cycle. The following conclusions are drawn:
1. At a cooling load of 210 kW and a cooling temperature of 303 K, the MVR compressor consumes only 14.227 kW, which is 77.04% less than a first-level energy-efficiency air-conditioning system. This demonstrates significant energy-saving potential for battery energy storage systems.
2. The system performance is sensitive to the cooling temperature and the compressor temperature rise. Increasing the cooling temperature from 283 K to 313 K reduces compressor power by about 10%, while decreasing the compressor temperature rise from 15 K to 10 K reduces power by about 32%. Therefore, optimizing these parameters can further improve the energy efficiency of battery energy storage systems cooling.
3. From an economic perspective, the MVR system has a lower initial cost (17,200 RMB vs. 24,200 RMB) and dramatically lower annual electricity cost. The payback period is only 2 months, making it a highly cost-effective solution for battery energy storage systems.
4. The indirect falling-film design ensures that batteries operate at atmospheric pressure, enhancing safety. The use of water as coolant reduces environmental and cost concerns compared to fluorinated liquids.
Future work will focus on experimental validation of the falling-film evaporator performance and the thermal uniformity across battery modules. The dynamic behavior under variable load conditions and the integration with battery energy storage systems management software also warrant investigation.