In this study, I present a comprehensive comparison of three common thermal management systems for lithium-ion battery energy storage systems: air cooling, indirect liquid cooling, and immersion liquid cooling. The rapid iteration of energy storage thermal management technology has seen a transition from air cooling to indirect liquid cooling in mainstream products, while immersion liquid cooling is moving from demonstration to application. However, there has been limited in-depth discussion regarding the economic advantages of immersion liquid cooling. Therefore, I aim to fill this gap by establishing an economic model that considers the characteristics of energy storage thermal management systems, including the impact of battery capacity degradation, to compare the costs, net present value, dynamic investment payback period, and internal rate of return. I also conduct a parameter sensitivity analysis on these indicators.
1. Introduction
Battery energy storage systems are crucial for integrating renewable energy sources and stabilizing power grids. Lithium-ion batteries, the most widely used electrochemical storage technology, generate significant heat during charging and discharging, especially at high rates. If this heat is not dissipated effectively, it can severely degrade battery performance and lifespan, potentially leading to thermal runaway. The primary thermal management technologies currently in the market include air cooling, indirect liquid cooling, and immersion liquid cooling. Air cooling uses air as the heat transfer medium, typically with air conditioners. Indirect liquid cooling uses a coolant in cold plates or pipes to exchange heat with the batteries. Immersion liquid cooling involves directly submerging the battery cells in a dielectric coolant. This study provides a multi-faceted comparison of these systems, focusing on key performance indicators and a detailed economic analysis.
2. Comparison of Key Performance Indicators
2.1 Weight and Volume Grouping Efficiency
Weight and volume grouping efficiency are key metrics for evaluating the integration density of battery energy storage systems. They are defined as the ratio of the total cell weight/volume to the total weight/volume of the energy storage container. Based on data from several commercial products, the average weight and volume grouping efficiencies for different thermal management systems are shown in the table below.
| Thermal Management System | Average Weight Grouping Efficiency (%) | Average Volume Grouping Efficiency (%) |
|---|---|---|
| Air Cooling | 45.90 | 15.92 |
| Indirect Liquid Cooling | 59.23 | 25.06 |
| Immersion Liquid Cooling | 44.19 | 26.45 |
The data indicates that immersion liquid cooling has a slightly lower weight grouping efficiency due to the added mass of the immersion tank and coolant. However, it achieves a higher volume grouping efficiency, which stems from its superior heat dissipation capability and the resulting higher battery packing density.
2.2 Temperature Control Capability
The temperature control capability of the three systems is a critical performance indicator. In specific test conditions, immersion liquid cooling can maintain the maximum temperature difference within a module at 2°C and an average temperature rise of 5°C. This is a significant improvement over indirect liquid cooling, which has a higher temperature difference, and air cooling, which performs poorly in both metrics.
Formally, we can define the thermal performance based on the maximum temperature rise ($\Delta T_{max}$) and the maximum temperature difference ($\Delta T_{diff}$) within the battery module.
$$
\Delta T_{max}^{Immersion} < \Delta T_{max}^{Indirect Liquid} < \Delta T_{max}^{Air Cooling}
$$
$$
\Delta T_{diff}^{Immersion} < \Delta T_{diff}^{Indirect Liquid} < \Delta T_{diff}^{Air Cooling}
$$
For battery energy storage systems, this superior thermal management translates directly to longer cycle life and better performance consistency.
3. Comparison of Operation and Maintenance (O&M)
The O&M requirements differ significantly between the three systems. A summary of regular maintenance items is provided in the table below.
| Maintenance Item | Air Cooling | Indirect Liquid Cooling | Immersion Liquid Cooling |
|---|---|---|---|
| Electrical Line Check | Visual inspection | Visual inspection | Visual inspection |
| Fan/Compressor Inspection | Check fan blades for smooth rotation | Check chiller fan blades | Check chiller fan blades |
| Condenser Cleaning | Check and clean condenser | Check and clean chiller condenser | Check and clean chiller condenser |
| Coolant Maintenance | Not applicable | Test coolant quality, replace if necessary; drain, vacuum, and refill | Test coolant quality, replace if necessary; drain and refill |
| Leak Detection | Not applicable | Use sensors to detect coolant components in the air | Use sensors to detect coolant components in the air |
For a battery pack fault, the replacement procedures also vary. The indirect and immersion liquid cooling systems require more steps, including draining coolant, removing piping, replacing the pack, reconnecting, and checking for leaks. While immersion cooling shares many steps with indirect liquid cooling, it does not always require a vacuum pump for the refill process, making it slightly simpler.
4. Economic Model for the Thermal Management System
4.1 Cost Model
To perform a comparative economic analysis, I established a comprehensive cost model for the thermal management system itself.
4.1.1 Equipment Cost ($E_p$)
The equipment cost is the sum of the costs of all individual components.
$$ E_p = \sum (p_e \times j) $$
Where $p_e$ is the unit price of a specific component, and $j$ is the number of that component required. The typical cost components for each system are summarized below.
Table: Cost Components for Different Thermal Management Systems
| System | Component 1 | Component 2 | Component 3 | Component 4 | Component 5 |
|---|---|---|---|---|---|
| Air Cooling | Container HVAC (1.4 ¥/W) | Galvanized Steel Ducts (40 ¥/m²) | Insulation Layer (10 ¥/m²) | Battery Pack Case (400 ¥/unit) | – |
| Indirect Liquid Cooling | Chiller (1.8 ¥/W) | Liquid Pipes (36.8 ¥/m) | Connectors (42 ¥/unit) | Water/Glycol Solution (7.3 ¥/kg) | Liquid Cold Plate (300 ¥/unit) |
| Immersion Liquid Cooling | Chiller (1.8 ¥/W) | Transformer Oil (15 ¥/L) | Liquid Pipes (36.8 ¥/m) | Connectors (42 ¥/unit) | Immersion Tank (400 ¥/unit) |
4.1.2 Operation and Maintenance Cost ($E_{o\&m}$)
The O&M cost comprises maintenance, coolant loss, and electricity consumption.
$$ E_{o\&m} = E_m + E_l + E_e $$
Where $E_m$ is the maintenance cost (typically a fixed percentage of equipment cost), $E_l$ is the coolant loss cost (only for liquid cooling systems), and $E_e$ is the electricity cost for the cooling equipment. Based on energy efficiency standards (GB 37479-2019 for air-cooled units, GB 19577-2015 for liquid chillers), the annual O&M costs for a 10 MWh system are calculated as 8.59 ¥/kWh for air cooling, 7.42 ¥/kWh for indirect liquid cooling, and 7.90 ¥/kWh for immersion liquid cooling.
4.1.3 Life Cycle Cost (LCC)
The LCC is the total cost over the system’s operational life, including installation and disposal.
$$ LCC = E_p + E_i + E_{o\&m} + E_d $$
Where $E_i$ is the installation cost and $E_d$ is the disposal cost. For a 20-year operational life, the LCC per kWh is calculated to be 208.13 ¥/kWh for air cooling, 199.25 ¥/kWh for indirect liquid cooling, and 204.21 ¥/kWh for immersion liquid cooling. The results show that while the initial equipment cost of immersion cooling is higher, its lower electricity consumption over the system’s life can make its LCC competitive.
5. Economic Analysis of the Battery Energy Storage System
To evaluate the overall economic viability of battery energy storage systems with different thermal management strategies, I established a system-level economic model.
5.1 System Cost Model
The total system cost ($C_{all}$) includes construction, operation electricity, O&M, and battery replacement costs.
$$ C_{all} = C_{self} + C_l + C_i + C_e + C_{o\&m} + C_b $$
Where $C_{self}$ is the self-financed portion, $C_l$ is the loan, $C_i$ is the interest, $C_e$ is the electricity cost for charging and thermal management, $C_{o\&m}$ is the system O&M cost, and $C_b$ is the cost of replacing the battery at the end of its life.
The battery replacement cost is a critical factor, as capacity degradation over time reduces system performance. The cost for a given year is calculated as:
$$ C_b = S \times (1 + i)^n \times c_b $$
Where $S$ is the system capacity, $i$ is the inflation rate, $n$ is the year of replacement, and $c_b$ is the unit battery price (0.36 ¥/Wh in this study).
5.2 System Revenue Model
The total revenue ($I_{all}$) is primarily derived from peak-valley price arbitrage, alongside other subsidies and incomes.
$$ I_{all} = I_s + I_d + I_o + I_c $$
Where $I_s$ is the government subsidy, $I_d$ is the income from peak-valley price arbitrage, $I_o$ is other operational income, and $I_c$ is the depreciation tax shield. The peak-valley arbitrage income is the main source, calculated as:
$$ I_d = S_n \times DOD \times (P_{sell} – P_{buy}) \times N $$
Where $S_n$ is the current capacity in year $n$, $DOD$ is the depth of discharge (90%), $P_{sell}$ is the selling price during peak hours (1.193 ¥/kWh), $P_{buy}$ is the buying price during valley hours (0.478 ¥/kWh), and $N$ is the number of cycles per year.
5.3 Evaluation Indicators
I used three primary indicators to evaluate the project.
5.3.1 Net Present Value (NPV)
$$ NPV = \sum_{t=0}^{N} \frac{(I_t – O_t)}{(1 + r)^t} $$
Where $I_t$ is total revenue in year $t$, $O_t$ is total cost in year $t$, $r$ is the discount rate (10%), and $N$ is the project lifetime (20 years).
5.3.2 Dynamic Investment Payback Period (DIPP)
$$ DIPP = n’ – 1 + \frac{(C_{investment} – \sum_{t=1}^{n’-1} (I_t – O_t))}{(I_{n’} – O_{n’})} $$
Where $n’$ is the year in which the cumulative net cash flow becomes positive.
5.3.3 Internal Rate of Return (IRR)
$$ 0 = \sum_{t=0}^{N} \frac{(I_t – O_t)}{(1 + IRR)^t} $$
The IRR is the discount rate that makes the NPV of the project equal to zero. A higher IRR indicates a more profitable project.
5.4 Results and Sensitivity Analysis
The economic analysis for a hypothetical 10 MWh battery energy storage system over 20 years is presented below. The EPC prices used are 1.17 ¥/Wh for air cooling, 1.30 ¥/Wh for indirect liquid cooling, and 1.40 ¥/Wh for immersion liquid cooling.
Table: Main Economic Indicators for Different Systems
| Indicator | Air Cooling | Indirect Liquid Cooling | Immersion Liquid Cooling |
|---|---|---|---|
| NPV (¥ million) | 1.05 | 2.80 | 4.80 |
| DIPP (years) | 15.2 | 10.5 | 8.2 |
| IRR (%) | 7.75 | 14.05 | 16.70 |
The data shows that the immersion liquid cooling system significantly outperforms the other two. Its NPV is 4.8 million ¥, which is 71% higher than indirect liquid cooling and 357% higher than air cooling. The payback period for immersion cooling is the shortest at 8.2 years, compared to 10.5 years for indirect liquid cooling and 15.2 years for air cooling. The IRR for immersion cooling is 16.70%, which is 3.2% higher than indirect liquid cooling and 9.1% higher than air cooling, making it a highly attractive investment.
5.4.1 Sensitivity Analysis of IRR
I performed a sensitivity analysis to understand how changes in key parameters affect the IRR. The analysis focused on the battery replacement price ($c_b$) and the electricity selling price ($P_{sell}$).
Table: Sensitivity of IRR to Battery Price (¥/Wh) and Electricity Sale Price (¥/kWh) for Air Cooling System
| 0.40 | 0.38 | 0.36 | 0.34 | 0.32 | |
|---|---|---|---|---|---|
| 1.05 | 1.80% | 2.70% | 3.61% | 4.53% | 5.47% |
| 1.10 | 5.80% | 6.77% | 7.75% | 8.73% | 9.72% |
| 1.15 | 10.22% | 11.26% | 12.30% | 13.35% | 14.40% |
Table: Sensitivity of IRR to Battery Price (¥/Wh) and Electricity Sale Price (¥/kWh) for Indirect Liquid Cooling System
| 0.40 | 0.38 | 0.36 | 0.34 | 0.32 | |
|---|---|---|---|---|---|
| 1.05 | 8.16% | 8.73% | 9.28% | 9.81% | 10.32% |
| 1.10 | 13.08% | 13.58% | 14.05% | 14.51% | 14.95% |
| 1.15 | 18.17% | 18.59% | 18.99% | 19.38% | 19.76% |
Table: Sensitivity of IRR to Battery Price (¥/Wh) and Electricity Sale Price (¥/kWh) for Immersion Liquid Cooling System
| 0.40 | 0.38 | 0.36 | 0.34 | 0.32 | |
|---|---|---|---|---|---|
| 1.05 | 11.98% | 12.25% | 12.52% | 12.79% | 13.05% |
| 1.10 | 16.20% | 16.45% | 16.70% | 16.95% | 17.19% |
| 1.15 | 20.65% | 20.88% | 21.10% | 21.32% | 21.53% |
The sensitivity analysis reveals that the IRR of all systems is more sensitive to the electricity sale price than to the battery price. However, the immersion cooling system is the least sensitive to both variables. The change in IRR per unit change in the electricity sale price ($\partial y / \partial x$) for immersion cooling is 5.65, compared to 7.6 for indirect liquid cooling and 12.33 for air cooling. Similarly, for battery price, the sensitivity for immersion cooling is 0.27, compared to 0.6 for indirect liquid cooling and 2.28 for air cooling. This indicates that immersion cooling provides a more stable return under market fluctuations.
6. Conclusion
This study provides a detailed comparative analysis of thermal management systems for battery energy storage systems, focusing on both technical performance and economic viability. I have developed a comprehensive economic model that incorporates the effects of battery degradation and thermal management system characteristics.
The key findings from my analysis are as follows:
- Technical Performance: Immersion liquid cooling offers the best temperature uniformity and lowest maximum temperature rise, outperforming both air and indirect liquid cooling. It also achieves a higher volume grouping efficiency, allowing for more compact battery energy storage systems.
- Operational and Maintenance: While liquid cooling systems have more complex maintenance procedures than air cooling, the benefits in terms of battery life and performance often outweigh the increased maintenance effort. Immersion cooling simplifies the coolant refill process compared to indirect liquid cooling.
- Life-Cycle Cost: Despite a higher initial cost, the immersion liquid cooling system achieves a competitive life-cycle cost, lower than air cooling, due to its superior energy efficiency and the extended lifespan it provides to the batteries.
- System Economics: For a 10 MWh battery energy storage system, the immersion liquid cooling system yields the highest Net Present Value (NPV), the shortest Dynamic Investment Payback Period (DIPP), and the highest Internal Rate of Return (IRR). The IRR for immersion cooling is 16.70%, which is a 3.2% absolute increase over indirect liquid cooling and a 9.1% increase over air cooling.
- Risk Mitigation: The immersion cooling system demonstrates the lowest sensitivity to fluctuations in battery prices and electricity sale prices. This makes it a more robust and less risky investment, particularly in volatile markets where the peak-valley price spread is subject to change.
In conclusion, for large-scale battery energy storage systems where long-term operational costs, battery longevity, and revenue stability are critical factors, immersion liquid cooling presents a superior economic and technical solution compared to air cooling and indirect liquid cooling. The initial higher investment is justified by the improved performance, lower lifetime energy costs, and enhanced financial returns over the system’s operational life.