EAGER: Renewables: A New Framework in Power System Protection with High Levels of Renewable Generation

风电和光伏发电在输电和配电系统中的高渗透率对使用现有方法维护保护系统的安全性和可靠性构成了威胁。从历史上看，几次重大的连锁停电都涉及保护继电保护的误操作或误配合。这样的停电表明，在紧张的系统条件下-并发高负荷需求、电路拓扑变化、设备停机和短路故障条件-期间继电保护不配合会导致配电网脆弱。过去，变量发电量明显低于现在的水平，因此不会导致继电保护不协调。由于预计未来风力和光伏水平只会增加，大规模可变发电可能会给现有保护系统带来额外的脆弱性。该项目的目标是开发新的工具，以便从根本上了解保护和控制系统应如何在高度可变的可再生能源存在的情况下运作。该项目旨在开发一个完全不同的框架和潜在的解决方案，该框架建立在基于模型的分布式继电保护智能、实时动态继电保护整定和随机优化的基础上。在设想的框架中，继电保护操作整定的计算被描述为具有本地继电保护智能的实时输入的随机优化问题，即预测电路模型。继电器收集的数据被输入到电路仿真，以便准确地预测继电器位置处的实时故障电流。然后，可以根据更新的生成配置文件实时调整设置。文献中提出的优化继电保护设置的方法一直受到限制，并且没有考虑可再生能源的随机性质。该项目将开发一种最优继电保护设置的概率公式，该公式自然地适应可再生能源的高渗透率带来的随机性。从预测电路模型以每个单独继电器的估计远程系统参数的形式提供输入。该优化允许每个继电器基于已知信息(系统结构和本地测量)和估计值(非本地条件)做出最佳可能的决策，以实时更新最优继电响应。这项工作将为纳入可再生能源引入的随机性的保护方法提供洞察力。这项工作将有助于增加可再生能源对电网的渗透率，从而有助于减少碳排放。所开发的方法将在硬件中实施，便于与业界讨论和向业界转让技术。

High penetration of wind and photovoltaic (PV) generation in transmission and distribution systems poses a threat to maintaining security and dependability of the protection system using existing approaches. Historically, several major cascading outages have involved mis-operation or mis-coordination of protective relays. Such outages demonstrate that relay mis-coordination during stressed system conditions---concurrent high load demand, changes in circuit topology, equipment outages, and short-circuit fault conditions---can result in a fragile distribution network. In the past, variable generation was at a significantly lower level than at present and thus did not contribute to relay mis-coordination. With levels of wind and PV projected only to increase in the future, large-scale variable generation can present an additional point of vulnerability to the existing protection system. The objective of this project is to pursue the development of new tools for the fundamental understanding of how protection and control systems should operate in the presence of highly variable renewable energy sources. The project aims to develop a radically different framework and potential solutions built upon on model-based distributed relay intelligence, real-time dynamic relay settings, and stochastic optimization.In the envisioned framework, the calculation of relay operating settings is formulated as a stochastic optimization problem with real-time inputs from local relay intelligence, i.e., the predictive circuit model. The data collected by a relay is input to circuit simulations in order to accurately predict real-time fault currents at the relay location. Settings can then be adapted in real time based on updated generation profiles. Methods proposed in the literature for optimal relay settings have been limited and did not account for the stochastic nature of renewable energy sources. This project will develop a probabilistic formulation of optimal relay settings that naturally adapts to the randomness introduced by the high penetration of renewable generation. Input is provided from the predictive circuit models in the form of estimated remote system parameters for each individual relay. The optimization allows each relay to make the best possible decision based on known information (system structure and local measurements) and estimated values (non-local conditions) to update the optimal relay response in real time. The work will provide insight for protection methodologies that incorporate randomness introduced by renewable energy sources. The work will contribute to allowing increased penetration of renewable generation sources to the grid, which contributes to reduction of carbon emissions. The developed methods will be implemented in hardware, facilitating discussions with and technology transfer to industry.