Limiting Energy Storage Cycles of Operation

Md Umar Hashmi, Ana Bušić2018 · DOI

TL;DR

This paper proposes a novel method to extend the lifespan of batteries used in energy storage systems by introducing a "friction" mechanism. This friction selectively filters out low-return arbitrage opportunities, thereby reducing the number of charge-discharge cycles and significantly improving battery longevity without a drastic loss in profit.

What Problem Does This Paper Address?

Batteries are a crucial component for integrating renewable energy sources (RES) into the power grid, helping to manage the inherent variability in power generation. However, batteries are expensive and have a finite lifespan, which is influenced by both calendar aging and the number of charge-discharge cycles they undergo. While energy storage systems can generate revenue through arbitrage (buying electricity when prices are low and selling when they are high), constantly cycling the battery for even small profits can accelerate its degradation, leading to premature end-of-life. This paper addresses the challenge of maximizing battery lifespan while still capitalizing on arbitrage opportunities, particularly by distinguishing between profitable and less profitable transactions.

Methods

The authors build upon their previous work on optimal arbitrage algorithms for energy storage. Their core contribution is the introduction of a "friction coefficient" into the control mechanism. This friction acts as a dead-band, effectively increasing the threshold for profitable charge-discharge cycles.

Here's how it works:

  • Modified Arbitrage Algorithm: The existing optimal arbitrage algorithm is modified to incorporate this friction.
  • Friction Coefficient: This coefficient, denoted as $N_{friction}$, is applied to the charging and discharging costs. A lower $N_{friction}$ value effectively raises the price difference required for a transaction to be considered profitable.
  • Selective Filtering: By increasing the "friction," the algorithm filters out arbitrage opportunities that offer only marginal returns. This means the battery will only cycle when the price difference is substantial enough to justify the wear and tear.
  • Adaptive Control: The paper suggests that the friction coefficient should ideally be tuned adaptively based on price volatility to balance arbitrage gains and battery operational life.

Datasets

The numerical results are based on real-time 15-minute locational marginal price (LMP) data from the New York Independent System Operator (NYISO) for December 21, 2016.

Key Insights and Ideas

  • Friction as a Lifespan Extender: The concept of "friction" is a clever way to manage battery cycling. By making it harder for the battery to initiate a charge or discharge, we force it to only act when the economic incentive is strong.
  • Trade-off Between Profit and Longevity: There's a clear trade-off between maximizing immediate arbitrage gains and extending battery life. The proposed friction mechanism allows users to navigate this trade-off by choosing an appropriate friction level.
  • Distinguishing Profitable Transactions: The friction mechanism inherently helps in distinguishing between truly profitable arbitrage opportunities and those that might lead to faster battery degradation for minimal gain.
  • Impact of Volatility: The paper implies that in highly volatile markets, a lower friction might be beneficial to capture more arbitrage opportunities, while in less volatile markets, a higher friction could be more appropriate to preserve battery life.

Key Results

The paper presents numerical results using NYISO data to demonstrate the effectiveness of the friction mechanism.

  • Optimal Friction: For the given dataset and battery parameters, an optimal friction coefficient of approximately 0.855 was found to achieve the target operational cycles per day ($I_{op}$) while balancing gains and cycles.
  • Cycle Reduction: At this optimal friction level, the number of cycles of operation decreased by approximately 36% compared to a scenario with no friction ($N_{friction}=1$).
  • Profit Reduction: Concurrently, the arbitrage gains were reduced by about 10%.
  • User-Defined Trade-off: For users prioritizing current gains, a friction of 0.928 resulted in only a 1.47% reduction in gains and a 11.7% reduction in cycles, showcasing the flexibility of the approach.
  • Shadow Price Convergence: As the friction coefficient is reduced, the shadow price of the transaction tends to converge to a single level, indicating less variation in the perceived value of electricity.

Limitations and Future Work

  • Assumed Temperature Control: The paper assumes that temperature is controlled and the state of charge (SoC) is regulated within an optimal band. In real-world scenarios, temperature fluctuations can significantly impact battery life.
  • Simplified Battery Model: The study uses a simplified model for battery life, focusing on cycle and calendar life. More complex degradation models could be incorporated.
  • Adaptive Tuning: While the paper suggests adaptive tuning of the friction coefficient, the specific algorithms for achieving this adaptation are not detailed.
  • Future Work: The authors suggest further research into adaptive tuning of the friction coefficient based on price volatility and exploring more sophisticated battery degradation models.

Summary updated

Abstract

International audience
BibTeX
@misc{hashmi2018limiting,
  author = {Md Umar Hashmi and Ana Bušić},
  title = {{Limiting Energy Storage Cycles of Operation}},
  year = {2018},
  doi = {10.1109/greentech.2018.00022},
}
Figure