This paper investigates the historical value of electricity storage from the perspective of the storage owner in day-ahead markets (DAM) across Europe. A technology-neutral formulation is used, where the storage is modelled based on its round-trip efficiency and storage duration. A mixed-integer linear program (MILP) is built to compute the perfect-foresight value of a price-taker storage from arbitrage, using historical hourly DAM prices in all the bidding zones of the EU-28 countries, Norway, Switzerland, and Turkey. Depending on the bidding zones, the DAM price data starts between 2000 and 2017, and spans to 2021. The model is solved for varying round-trip efficiencies (50% to 100%) and storage durations (1 to 10 h) for every bidding zone and every year in the dataset. The results reveal significant variations in storage value from arbitrage, both geographically and temporally, with round-trip efficiency having a major impact on arbitrage value and storage duration having very low marginal value beyond 4 to 6 h. Additionally, the paper investigates the impact of variable grid fees on arbitrage value, using the case of Belgium, where fees depend on storage system size. The initial MILP is therefore augmented to account for the complex dependencies between storage size and the resulting grid fees. The augmented MILP shows that grid fees can decrease storage arbitrage value by 20% to 50%, and that they can also dramatically decrease storage participation in DAMs.

本文从欧洲日前市场 (DAM) 中储能所有者的角度研究了储能的历史价值。使用技术中立的公式，其中存储基于其往返效率和存储持续时间建模。使用欧盟 28 个国家、挪威、瑞士和土耳其的所有投标区的历史每小时 DAM 价格，构建了一个混合整数线性程序 (MILP) 来计算套利中价格接受者存储的完美预见值. 根据投标区域，DAM 价格数据从 2000 年到 2017 年开始，一直到 2021 年。该模型针对每次投标的不同往返效率（50% 到 100%）和存储持续时间（1 到 10 小时）进行求解区域和数据集中的每年。结果揭示了套利的存储价值的显着变化，在地理和时间上，往返效率对套利价值和存储持续时间有重大影响，超过 4 至 6 小时的边际价值非常低。此外，本文以比利时为例，调查了可变电网费用对套利价值的影响，比利时的费用取决于储能系统的规模。因此，增加了初始 MILP 以解决存储大小和由此产生的网格费用之间的复杂依赖关系。增强的 MILP 表明，网格费用可以将存储套利价值降低 20% 到 50%，并且它们还可以显着减少 DAM 中的存储参与。该论文以比利时为例，调查了可变电网费用对套利价值的影响，比利时的费用取决于储能系统的规模。因此，增加了初始 MILP 以解决存储大小和由此产生的网格费用之间的复杂依赖关系。增强的 MILP 表明，网格费用可以将存储套利价值降低 20% 到 50%，并且它们还可以显着减少 DAM 中的存储参与。该论文以比利时为例，调查了可变电网费用对套利价值的影响，比利时的费用取决于储能系统的规模。因此，增加了初始 MILP 以解决存储大小和由此产生的网格费用之间的复杂依赖关系。增强的 MILP 表明，网格费用可以将存储套利价值降低 20% 到 50%，并且它们还可以显着减少 DAM 中的存储参与。