CAREER: Risk-Aware Power System Operations with Significant Wind Power Penetration

风力发电已以快速增长的速度被纳入国家的大容量电力系统。由于风力发电的间歇性、波动性和不确定性，给电力系统的运行和规划带来了巨大的挑战。该项目旨在解决风电整合中的关键挑战，包括储备采购、风电坡道和非自愿弃风。本项目的研究将产生以下成果：1)风电坡道风险评估和量化的系统方法；2)极端天气诱发的大型风力发电坡道预警系统；3)从电网发电和输电资源中获取储备的成本效益操作协议。研究成果将被整合到开发新的多学科课程中，以丰富国家风能研究所的风力工程课程，并通过设计和实验项目吸引本科生在专门的风力研究实验室进行研究，该实验室将作为项目的成果而建立。预计项目成果将通过增强运营商对风力发电坡道的态势感知，以及通过提高大容量电力系统和批发电力市场的可靠性、安全性和效率，影响电力系统的运行。综合研究和教育活动将有助于培训合格的工程师和研究人员，他们可以为繁荣和可持续的风能产业做出贡献。该项目研究寻求在风力发电坡道风险评估和减少弃风方面取得根本性突破，以便在大型电力系统中有效利用增加的风力发电能力。大型风力坡道表现出由广义帕累托分布决定的尾部行为这一重要观察结果，促使我们开发了一种量化风力坡道风险的系统方法，并在此基础上严格确定足够的储量。沿着不同的路径，使用Mesonet和分散的风力发电场的真实数据的初步研究表明，Mesonet的测量结果确实包含由极端天气事件（锋面，雷暴，结冰事件等）引起的大型风力坡道的关键特征，最先进的风力预测系统可能无法捕获这些特征。有了这一见解，基于mesonet的早期报警系统将被设计出来，以提高电力系统运营商对大型风力坡道的风险意识。此外，将从限制拥堵风险的角度发展一种创新的线路转移裕度概念，通过抑制拥堵风险和提高储备的可交付性，有助于减少非自愿弃风。利用节点风力发电的基本统计信息，结合电网的潮流分配系数，可以很容易地预先计算出这些线路转移余量。研究采用了几种创新和非传统的方法，包括利用极值理论对风力坡道进行风险量化，通过网络数据分析对风力坡道事件进行检测，以及从图论角度对动态储备分区进行探索性研究。

Wind power has been integrated into the nation's bulk power systems at a rapidly increasing pace. Due to its intermittency, volatility, and uncertainty, wind power generation has posed grand challenges for power system operations and planning. This project aims at addressing key challenges in wind power integration, including reserve procurement, wind power ramps, and involuntary wind power curtailment. The research of this project will produce the following: 1) a systematic approach for risk assessment and quantification of wind power ramps; 2) an early alarm system for large wind power ramps induced by extreme weather events; and 3) cost-effective operational protocols for acquiring reserves from generation and transmission resources of power networks. The research outcomes will be integrated to develop new multidisciplinary courses to enrich the curriculum of wind engineering program at the National Wind Institute, and to engage undergraduate students in research through design and experimental projects at a dedicated wind power research laboratory that is to be established as an outcome of the project. The project outcomes are expected to impact power system operations by enhancing the operator's situational awareness of wind power ramps, and by improving the reliability, security, and efficiency of bulk power systems and the wholesale electricity market. The integrated research and educational activities will contribute to training qualified engineers and researchers who can contribute to a thriving and sustainable wind energy industry.The project research seeks fundamental breakthroughs in wind power ramp risk assessment and curtailment reduction to enable efficient utilization of increased wind power capacity in bulk power systems. Motivated by a key observation that large wind power ramps exhibit tail behaviors dictated by generalized Pareto distributions, a systematic method for quantifying wind power ramp risk is developed, based on which the adequate amount of reserves can be determined in a rigorous manner. Along a different path, preliminary studies using real-world data from Mesonet and dispersed wind farms reveal that Mesonet measurements indeed contain critical signatures of large wind power ramps induced by extreme weather events (fronts, thunderstorms, icing events, etc.), which state-of-the-art wind power forecasting systems may fail to capture. With this insight, a Mesonet-based early alarm system will be designed to enhance the power system operator's risk awareness with respect to large wind power ramps. Further, an innovative concept of line transfer margin will be developed from a congestion risk-limiting viewpoint, which facilitates reduction of involuntary wind power curtailment by suppressing congestion risk and by improving the deliverability of reserves. These line transfer margins can be easily pre-computed by using basic statistical information on nodal wind power generation together with the flow distribution factors of power networks. The research encompasses several innovative and nontraditional approaches, including risk quantification of wind power ramps using extreme value theory, wind power ramp events detection through networked data analytics, and exploratory study of dynamic reserve zoning from a graph-theoretic perspective.