Integrated risk measurement and control for stochastic energy trading of a wind storage system in electricity markets

Dongliang Xiao, Haoyong Chen, Wei‐Jun Cai, Chun Wei, Zhendong ZhaoProtection 2023 · DOI

TL;DR

This paper introduces a novel integrated risk management framework for wind-plus-storage systems trading in volatile electricity markets. By simultaneously optimizing for three distinct risk metrics—Value at Risk (VaR), Conditional Value at Risk (CVaR), and Shortfall Probability (SP)—the approach allows power producers to maximize profits while precisely controlling the likelihood and severity of financial losses.

What Problem Does This Paper Address?

As renewable energy penetration grows, electricity markets face increased price volatility. Wind power producers face a "double whammy" of uncertainty: they don't know exactly how much energy they will produce (intermittency), and they don't know what the market price will be when they sell it. While adding battery energy storage systems (BESS) helps mitigate these risks through arbitrage and deviation reduction, deciding how to trade remains a complex challenge.

Most existing research focuses on a single risk metric, usually CVaR, which looks at the average of the worst-case scenarios. However, different decision-makers have different "risk flavors." Some care about the probability of failing to meet a target (Shortfall Probability), while others care about a specific loss boundary (Value at Risk). This paper fills the gap by providing a unified mathematical framework that handles all three simultaneously, offering a more flexible "control knob" for risk-averse operators.

Methods

The authors employ a Scenario-Based Risk-Aware Stochastic Optimization model. Key components include:

  • Uncertainty Modeling: They use SARIMA (Seasonal Autoregressive Integrated Moving Average) models to generate realistic scenarios for wind power production and electricity prices in both Day-Ahead (DA) and Real-Time (RT) markets.
  • Integrated Risk Measurement (IRM): This is the core innovation. The objective function is a weighted sum of:
  • Expected Profit: The "risk-neutral" goal.
  • Shortfall Probability (SP): The probability that profit falls below a specific benchmark.
  • Value at Risk (VaR): The maximum potential loss at a specific confidence level (the "boundary").
  • Conditional Value at Risk (CVaR): The expected profit within the worst-case tail of the distribution.
  • Optimization: The model is formulated as a mixed-integer linear programming (MILP) problem and solved using the YALMIP toolbox with the Gurobi solver in MATLAB.

Datasets

The study validates its approach using realistic market data:

  • Wind Power: Historical data from the National Renewable Energy Laboratory (NREL).
  • Electricity Prices: Historical Day-Ahead and Real-Time price data from the PJM (Pennsylvania-Jersey-Maryland) interconnection, specifically the IMO trading hub.
  • System Specs: A simulated 16 MW wind farm paired with a 6 MWh / 3 MW battery storage system.

Key Insights and Ideas

  • The "Price Spread" Trigger: Risk management is most critical when the price difference between Day-Ahead and Real-Time markets is small. When the spread is large, the profit motive is so strong that risk-neutral and risk-averse strategies converge.
  • Arbitrage Logic: The system intelligently learns to sell in the DA market during high-price windows (e.g., 11:00–14:00) and "buy back" energy in the RT market to cover deviations, using the battery to buffer the timing.
  • Diminishing Returns of Risk Aversion: Increasing the risk-aversion parameter ($\beta_r$) beyond 0.4 yields very little additional risk reduction but starts to eat significantly into expected profits. The authors suggest 0.3 to 0.4 as a "sweet spot."
  • Battery Stabilization: Increasing battery capacity doesn't just increase profit; it significantly stabilizes the tail risks (VaR and CVaR), though the improvement in Shortfall Probability is less linear due to the stochastic nature of extreme weather/price events.

Key Results

  • Risk vs. Reward: Compared to a standard risk-neutral strategy, the integrated approach increased the minimum profit by 75.67%.
  • Minimal Sacrifice: This massive boost in safety came at a cost of only a 1.58% decrease in total expected profit, representing a highly efficient trade-off.
  • Metric Improvement: The integrated strategy successfully reduced the Shortfall Probability by 33.33% while simultaneously improving VaR and CVaR by 57.63% and 35.13%, respectively, compared to the risk-neutral baseline.

Limitations and Future Work

  • Resource Diversity: The current model only considers wind and batteries. The authors suggest that future iterations should incorporate other flexible assets like Electric Vehicles (EVs), gas-fired units, or hydrogen storage.
  • Market Complexity: The study focuses on energy markets. Future work could expand the framework to include ancillary services (like frequency regulation) or long-term bilateral contracts, which have different risk profiles.
  • Computational Load: Scenario-based optimization can become computationally expensive as the number of scenarios grows. Exploring more efficient uncertainty sets (like robust optimization) could be a promising direction.

Summary updated

Abstract

Abstract To facilitate wind energy use and avoid low returns, or even losses in extreme cases, this paper proposes an integrated risk measurement and control approach to jointly manage multiple statistical properties of the expected profit distribution for a wind storage system. First, a risk-averse stochastic decision-making framework and multi-type risk measurements, including the conditional value at Risk (CVaR), value at risk (VaR) and shortfall probability (SP), are described in detail. To satisfy the various needs of multi-type risk-averse decision makers, integrated risk measurement and control approaches are then proposed by jointly considering the expected, boundary and probability values of the extreme results. These are managed using CVaR, VaR and SP, respectively. Finally, the effectiveness of the proposed risk control strategy is verified by conducting case studies with realistic market data, and the results of different risk control strategies are analyzed in depth. The impacts of the risk parameters of the decision maker, the energy capacity of the battery storage and the price difference between the day-ahead and real-time markets on the expected profits and risks are investigated in detail.
BibTeX
@inproceedings{xiao2023integrated,
  author = {Dongliang Xiao and Haoyong Chen and Wei‐Jun Cai and Chun Wei and Zhendong Zhao},
  title = {{Integrated risk measurement and control for stochastic energy trading of a wind storage system in electricity markets}},
  booktitle = {Protection},
  year = {2023},
  doi = {10.1186/s41601-023-00329-3},
}
Figure