Integrated Energy Exchange Scheduling for Multimicrogrid System With Electric Vehicles

Dai Wang, Xiaohong Guan, Jiang Wu, Pan Li, Peng Zan, Hui XuIEEE Trans. Smart Grid 2016 · DOI

TL;DR

This paper proposes a two-stage scheduling strategy for multi-microgrid systems that integrates electric vehicles (EVs) as flexible battery storage. By developing an "updated price signal" derived from dual variables, the approach enables decentralized scheduling for individual microgrids, effectively reducing peak power exchange with the main grid and lowering electricity costs.

What Problem Does This Paper Address?

The increasing integration of renewable energy sources and electric vehicles into power grids presents challenges for grid stability and cost management. Specifically, when multiple microgrids operate independently, their tendency to charge during off-peak hours and discharge during peak hours can lead to significant spikes in power exchange with the main grid, potentially exceeding safe limits. Furthermore, the dynamic charging and discharging behaviors of EVs, influenced by user driving patterns, introduce variability into the aggregate energy storage capacity of microgrids. This paper aims to address these issues by developing a coordinated scheduling framework that optimizes energy exchange for a multi-microgrid system, considering EV integration and aiming to mitigate peak power demand and reduce overall electricity costs.

Methods

The paper introduces a two-stage integrated energy exchange scheduling strategy:

  1. MGCC Scheduling Stage (Centralized): The Microgrid Central Controller (MGCC) treats all EVs as an aggregate battery storage. This stage aims to minimize the total electricity cost of the multi-microgrid system while limiting the total power exchange with the main grid within a safe range. It employs a scenario tree method to handle the stochastic nature of wind power and EV driving patterns.
  2. EV Aggregator Scheduling Stage (Decentralized): Once the optimal total charging/discharging power is determined by the MGCC, the EV aggregator allocates this power to individual EVs. This stage also focuses on minimizing transitions between charging and discharging states to prolong battery life.

A key innovation is the development of a decentralized scheduling strategy based on an "updated price signal." This signal is derived from the dual variables of the centralized optimization problem, which represent the marginal cost of power exchange constraints. By broadcasting this updated price signal to each microgrid, the system allows for decentralized decision-making. Each microgrid then optimizes its local energy management based on this dynamic price, effectively coordinating its actions without direct central control over individual devices. This approach aims to achieve the benefits of centralized optimization (peak shaving, cost reduction) while enabling decentralized operation.

Datasets

The paper utilizes real-world driving data from the NHTS2009 dataset to model EV driving patterns (arrival time, departure time, number of daily trips, and distance of each trip). Forecasted wind speed data for Tianjin Eco-city is also used.

Key Insights and Ideas

  • EVs as Aggregate Batteries: EVs, when coordinated, can act as a significant and flexible distributed energy storage resource for microgrids.
  • The "Updated Price Signal": Deriving a dynamic price signal from dual variables of a centralized optimization problem is an effective way to translate global objectives (like peak shaving) into local decision-making for decentralized control.
  • Decentralization for Scalability and Fairness: While centralized control can optimize the entire system, a decentralized approach based on price signals offers better scalability and potentially fairer cost distribution among microgrids.
  • Balancing Cost and Battery Life: The EV aggregator stage explicitly considers minimizing battery state transitions, recognizing that frequent charging/discharging cycles can degrade battery life and incur costs.
  • Stochasticity Management: The use of scenario tree methods and subsequent scenario reduction is crucial for handling the inherent uncertainties in renewable energy generation and EV behavior.

Key Results

  • Peak Power Reduction: The proposed coordinated strategy reduced the peak power exchange with the main grid by 10.18% compared to an uncoordinated strategy. In the worst-case scenario, the peak power exchange was reduced by 18.77% compared to a decentralized strategy using only the original TOU price.
  • Cost Reduction: The total electricity cost was reduced by 12.97% compared to scheduling without EV coordination.
  • Reduced Standard Deviation: The standard deviation of power exchange profiles was reduced by 16.49%, indicating a more stable and predictable power flow.
  • Battery Cycle Life: The EV aggregator's allocation strategy aims to minimize unnecessary charging/discharging cycles, potentially saving significant costs related to battery degradation.
  • Decentralized vs. Centralized: The decentralized strategy with the updated price signal achieved similar peak power reduction to the centralized strategy but with a slightly higher average electricity cost (13452 CNY vs. 13362 CNY). This trade-off is attributed to the price signal's robustness adjustments.

Limitations and Future Work

  • AC Network Constraints: The current model primarily uses a DC power flow approximation. The authors acknowledge that extending the model to incorporate detailed AC operational constraints and system dynamics is a direction for future work.
  • Battery Degradation Model: The paper simplifies battery degradation to be primarily dependent on the number of charging/discharging cycles. More complex models considering factors like ambient temperature and depth of discharge could be explored.
  • Computational Burden: While scenario reduction techniques are used, managing a large number of scenarios for stochastic optimization can still be computationally intensive.
  • Economic Fairness: The paper notes that while the decentralized strategy is based on price signals, it doesn't explicitly guarantee economic fairness for each microgrid, which is a motivation for the price-based approach. Future work could further investigate this aspect.

Summary updated

BibTeX
@article{wang2016integrated,
  author = {Dai Wang and Xiaohong Guan and Jiang Wu and Pan Li and Peng Zan and Hui Xu},
  title = {{Integrated Energy Exchange Scheduling for Multimicrogrid System With Electric Vehicles}},
  journal = {IEEE Trans. Smart Grid},
  year = {2016},
  doi = {10.1109/tsg.2015.2438852},
}
Figure