# Energy Storage Price Arbitrage via Opportunity Value Function Prediction

arXiv: https://arxiv.org/abs/2211.07797 · HTML: https://mlsanity.com/paper/2211.07797/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 20 references that are in the map.
- First posted on arXiv 14 Nov 2022.

## Builds on (most cited first, 6 of 6)

- [Energy Storage Arbitrage in Real-Time Markets via Reinforcement Learning](https://mlsanity.com/paper/1711.03127/) (2017)
- [Arbitraging Variable Efficiency Energy Storage using Analytical Stochastic Dynamic Programming](https://mlsanity.com/paper/2108.06000/) (2021)
- [Impact of Local Transmission Congestion on Energy Storage Arbitrage Opportunities](https://mlsanity.com/paper/1702.04816/) (2017)
- [Operational Valuation for Energy Storage under Multi-stage Price Uncertainties](https://mlsanity.com/paper/1910.09149/) (2019)
- [Impact of Bidding and Dispatch Models over Energy Storage Utilization in Bulk Power Systems](https://mlsanity.com/paper/2201.03421/) (2022)
- [A Data-driven Nonlinear Recharge Controller for Energy Storage in Frequency Regulation](https://mlsanity.com/paper/2101.04858/) (2021)

## Built on by (most cited first, 2 of 2)

- [Distributional Reinforcement Learning-based Energy Arbitrage Strategies in Imbalance Settlement Mechanism](https://mlsanity.com/paper/2401.00015/) (2023)
- [Transferable Energy Storage Bidder](https://mlsanity.com/paper/2301.01233/) (2023)

## Abstract (arXiv)

This paper proposes a novel energy storage price arbitrage algorithm combining supervised learning with dynamic programming. The proposed approach uses a neural network to directly predicts the opportunity cost at different energy storage state-of-charge levels, and then input the predicted opportunity cost into a model-based arbitrage control algorithm for optimal decisions. We generate the historical optimal opportunity value function using price data and a dynamic programming algorithm, then use it as the ground truth and historical price as predictors to train the opportunity value function prediction model. Our method achieves 65% to 90% profit compared to perfect foresight in case studies using different energy storage models and price data from New York State, which significantly outperforms existing model-based and learning-based methods. While guaranteeing high profitability, the algorithm is also light-weighted and can be trained and implemented with minimal computational cost. Our results also show that the learned prediction model has excellent transferability. The prediction model trained using price data from one region also provides good arbitrage results when tested over other regions.