In modern power systems, energy storage cells play a pivotal role in enhancing grid stability, integrating renewable energy sources, and optimizing economic returns. The ability of battery energy storage systems (BESS) to provide diverse services, such as frequency regulation and energy arbitrage, has garnered significant attention. However, these applications differ substantially in terms of investment returns, technical performance, and cycle life. For instance, frequency regulation involves rapid, shallow charge-discharge cycles, while energy arbitrage requires deeper, less frequent cycles. This divergence poses a challenge for maximizing the lifespan and profitability of energy storage cells. To address this, we propose a comprehensive whole-life-cycle planning method that allocates the capacity of energy storage cells across multiple application scenarios, thereby optimizing overall benefits. The core idea is to deploy energy storage cells in the auxiliary service market for frequency regulation during their first life cycle phase, when the cells are new and have high efficiency. As the capacity degrades over time, the cells are transitioned to the energy arbitrage market for load shifting in the second life cycle phase. This approach not only extends the useful life of energy storage cells but also enhances economic gains by leveraging their residual capacity effectively.
The planning framework for energy storage cells involves a detailed mathematical model that considers revenues, costs, and degradation effects across both life cycle phases. In the first phase, energy storage cells participate in frequency regulation services, where they respond to grid frequency deviations within seconds to minutes. The power output is determined based on the frequency error and the predefined P-f characteristics, with a dead band to avoid unnecessary cycling. The revenue from this service is derived from a fixed capacity payment, while costs include operation and maintenance (O&M), degradation due to cycling, and penalties for unmet regulation demands. In the second phase, the energy storage cells are used for energy arbitrage, charging during low-price periods and discharging during high-price periods to capitalize on electricity price differentials. The revenue here is based on the actual energy traded, and costs include O&M and degradation. The objective function maximizes the net present value of total profits over the entire life cycle, discounted at an appropriate rate, while accounting for the initial investment in energy storage cells. This holistic model ensures that the planning of energy storage cells balances short-term revenues with long-term durability, making it a robust strategy for various grid applications.
To illustrate the operational principles, consider the frequency regulation service provided by energy storage cells. The output power is governed by a piecewise linear function of frequency deviation. Let $$ f_{\text{rated}} $$ be the nominal grid frequency (e.g., 50 Hz or 60 Hz), and define a dead band of $$ \pm 0.033 $$ Hz to prevent excessive cycling. The frequency regulation coefficient $$ \sigma $$ is calculated as: $$ \sigma = -\frac{\Delta f / f_{\text{rated}}}{\Delta P / P_{\text{bat,max}}^2} $$ where $$ \Delta f $$ is the frequency deviation, $$ \Delta P $$ is the power adjustment, and $$ P_{\text{bat,max}} $$ is the maximum power capacity of the energy storage cells allocated for ancillary services. For a given scenario $$ \xi $$ and time interval $$ t $$, the charging or discharging power $$ P^{\xi}_{\text{ch/dis},i,t} $$ of the i-th energy storage cell is: $$ P^{\xi}_{\text{ch/dis},i,t} = \pm P^{i}_{\text{bat,max}} \left[ \Delta f(t) / f_{\text{rated}} \right] / \sigma $$ This ensures that the energy storage cells respond proportionally to frequency errors, thereby supporting grid stability. The revenue from frequency regulation is: $$ \text{Profit}_1 = C_{\text{reserve}} \times C^{i}_{\text{bat,max}} $$ where $$ C_{\text{reserve}} $$ is the market price for reserve capacity (e.g., in $/MW/year), and $$ C^{i}_{\text{bat,max}} $$ is the energy capacity of the energy storage cells in the first life cycle phase.
In the energy arbitrage market, energy storage cells operate on a longer time scale, typically hours, to exploit price variations. The revenue is computed as: $$ \text{Profit}_2 = \sum \left( \lambda_{\text{peak}}(h) \times P^{\zeta}_{\text{dis},i,h} – \lambda_{\text{off-peak}}(h) \times P^{\zeta}_{\text{ch},i,h} \right) $$ where $$ \lambda_{\text{peak}} $$ and $$ \lambda_{\text{off-peak}} $$ are the electricity prices during peak and off-peak periods, respectively, and $$ P^{\zeta}_{\text{dis},i,h} $$ and $$ P^{\zeta}_{\text{ch},i,h} $$ are the discharge and charge powers in scenario $$ \zeta $$ at time $$ h $$. The degradation of energy storage cells is modeled using an effective cycle aging cost, which depends on the number of equivalent full cycles and the rated cycle life. For the first life cycle phase, the aging cost $$ C_{\text{age}} $$ is: $$ C_{\text{age}} = (C_{\text{inv}} / \Gamma_R) \times \sum_{\xi} \sum_{t} \Gamma_{\text{eff}}(t) $$ where $$ C_{\text{inv}} $$ is the initial investment cost, $$ \Gamma_R $$ is the rated cycle life of the energy storage cells, and $$ \Gamma_{\text{eff}}(t) $$ is the effective aging per time interval. Similarly, for the second phase, the aging cost $$ C’_{\text{age}} $$ is: $$ C’_{\text{age}} = (C_{\text{inv}} / \Gamma’_R) \times \sum_{\zeta} \sum_{h} \Gamma’_{\text{eff}}(h) $$ where $$ \Gamma’_R $$ accounts for the reduced cycle life in the second phase due to prior degradation.
The overall objective function for the whole-life-cycle planning of energy storage cells combines profits and costs from both phases, minus the initial investment. It is expressed as: $$ \max \left\{ \left[ \sum_{y}^{Y_{\text{1st}}} \left( \text{Profit}_1 – \frac{365}{S} \sum_{\xi}^{S} \sum_{t}^{T} \left( C_{\text{opm}} + C_{\text{age}} + C_{\text{penalty}} \right) \right) \right] + \left[ \sum_{y=Y_{\text{1st}}}^{Y_{\text{2nd}}} \left( \frac{365}{S’} \sum_{\zeta}^{S’} \left( \text{Profit}_2 – \sum_{h}^{H} \left( C’_{\text{opm}} + C’_{\text{age}} \right) \right) \right) \right] – C_{\text{inv}} \right\} $$ where $$ C_{\text{opm}} = \sum (P^{i}_{\text{bat,max}} C_{O,\text{BESS}}) $$ is the O&M cost, $$ C_{\text{penalty}} = \frac{365}{S} \sum_{\xi} \rho \times \Delta P^{\xi}_{\text{dev}}(t) \times \Delta t $$ is the penalty for regulation deviations, and $$ C_{\text{inv}} = \frac{d(1+d)^{(Y_{\text{1st}}+Y_{\text{2nd}})}}{(1+d)^{(Y_{\text{1st}}+Y_{\text{2nd}})} + 1 – 1} \sum C_E C^{i}_{\text{bat,max}} $$ is the annualized investment cost with discount rate $$ d $$. This formulation ensures that the planning of energy storage cells accounts for temporal variations, multiple scenarios, and the trade-off between immediate revenues and long-term degradation.
To validate the proposed planning method for energy storage cells, we conducted simulations on a modified IEEE 33-node distribution system. This system includes distributed generation, such as wind farms at nodes 19 and 32, to represent high renewable penetration. The load data and feeder parameters are based on standard test cases, with hourly wind speed and load demand derived from 10-year historical averages. The energy storage cells are assumed to be lithium-ion batteries due to their favorable performance characteristics, including high efficiency and cycle life. Key parameters for the energy storage cells include: state-of-charge (SOC) limits of 20% to 80%, initial SOC of 50%, charge-discharge efficiency of 75%, and a rated cycle life of 4580 cycles. The cost parameters are set as follows: energy cost $$ C_E = 325 $/kWh $$, O&M cost $$ C_{O,\text{BESS}} = 20 $/day $$, and ancillary service price $$ C_{\text{reserve}} = 120 $/kW/year $$. The discount rate is 5%, and the planning horizon is 15 years. Multiple scenarios (S=4 for frequency regulation, S’=4 for energy arbitrage) are used to capture uncertainties in wind generation and load demand.
We compared four cases to demonstrate the effectiveness of the whole-life-cycle planning approach for energy storage cells. Case 1 involves using energy storage cells solely for energy arbitrage (load shifting), with one full cycle per day. Case 2 dedicates energy storage cells only to frequency regulation, with high cycling frequency. Case 3 employs energy storage cells for both services simultaneously without phase separation. Case 4 implements the proposed method, where energy storage cells are allocated to frequency regulation in the first phase and energy arbitrage in the second phase, with no restrictions on capacity allocation ratio. The results, summarized in the table below, highlight the trade-offs between lifetime, costs, and revenues. For instance, in Case 1, the energy storage cells have a longer lifespan (12.5 years) but lower overall profit due to limited revenue streams. In Case 2, the lifespan is short (4.35 years) but revenues are high initially; however, accelerated degradation reduces long-term benefits. Case 3 shows the shortest lifespan (2.16 years) due to simultaneous high-stress services, yet it achieves high annual revenues. In contrast, Case 4 strikes a balance, with a lifespan of approximately 10.7 years and the highest total profit, as it leverages the strengths of both applications while mitigating degradation effects. This underscores the importance of phased deployment in maximizing the value of energy storage cells over their entire life cycle.
| Case | Application Scenarios | Node Locations | BESS Capacity (MWh) | Lifetime (years) | Investment Cost ($) | O&M Cost ($) | Frequency Regulation Revenue ($) | Energy Arbitrage Revenue ($) | Total Life-Cycle Profit ($) |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Energy Arbitrage Only | 31, 18 | 3.15, 1.46 | 12.5 | 813,330 | 28,904 | 0 | 2,300,088 | 1,380,864 |
| 2 | Frequency Regulation Only | 29, 14 | 3.09, 1.52 | 4.35 | 1,210,299 | 10,191 | 2,563,531 | 0 | 1,343,040 |
| 3 | Simultaneous Services | 30, 13 | 3.26, 1.32 | 2.16 | 1,339,614 | 4,946 | 1,246,161 | 202,407 | 1,784,347 |
| 4 | Proposed Phased Approach | 30, 14 | 2.77, 1.84 | 10.7 | 889,049 | 19,958 | 589,407 | 2,438,818 | 2,119,217 |
The simulation results clearly indicate that the proposed whole-life-cycle planning method for energy storage cells outperforms single-service strategies in terms of both economic returns and operational longevity. By sequentially allocating energy storage cells to high-revenue, high-degradation services followed by lower-stress applications, the method achieves a superlinear gain in net benefits. For example, compared to Case 1, Case 4 increases profit by 34.8%; compared to Case 2, by 36.6%; and compared to Case 3, by 15.8%. This is attributed to the optimal utilization of the energy storage cells’ capacity across different phases, which reduces the effective degradation rate per revenue unit. Additionally, the method enhances grid reliability by ensuring that energy storage cells continue to provide valuable services even after partial capacity fade. The use of multiple scenarios in the optimization further robustifies the plan against uncertainties in renewable generation and market prices, making it a scalable solution for various distribution networks.
In conclusion, the whole-life-cycle planning of energy storage cells for multiple application scenarios represents a significant advancement in energy storage management. It addresses the inherent trade-offs between profitability and durability by dynamically adapting the usage strategy based on the state of health of the energy storage cells. This approach not only maximizes the economic value of energy storage cells but also contributes to grid stability and renewable energy integration. Future work could explore real-time adaptation mechanisms, integration with other storage technologies, and policy incentives to further enhance the feasibility of such planning methods. Overall, this research provides a foundational framework for stakeholders to deploy energy storage cells in a cost-effective and sustainable manner, ensuring long-term benefits for the entire power system.