Collaborative Research: Robust Asset-and-User-Aware Dispatch of the Power Distribution Grid during Extreme Temperatures

极端温度可能会将各种电网组件推向其运行极限。当气温超过一定的阈值时，大多数发电资源和电力系统部件的可用容量会受到负面影响。不足为奇的是，温度导致的可用发电和输电能力的下降通常与对系统的电力需求增加相一致，这主要归因于空调(A/C)系统使用率的增加。忽略温度对各种电网资产的影响可能会导致这些资产过载，导致寿命缩短和组件过早故障。因此，至关重要的是将环境温度的影响纳入电网运行，以防止部件受到压力，并避免在极端温度期间无法满足需求而可能导致的停电。这个问题变得越来越重要，因为气候模型预测热浪的持续时间和频率都会增加。在极端温度条件下停电不仅是一种不便，因为它还可能影响其他关键基础设施的供应，如水卫生厂、运输系统、医院和其他紧急护理单元。在这个项目中，研究人员将寻求一种可能的解决方案，包括设计一种方法，用于主动调度暴露在极端环境温度下的配电系统中的能源。电力公用事业公司传统上通过两种方法来解决手头的问题：为各种组件定义动态热额定(DTR)，以根据环境温度调整其可用容量；最近，提供激励需求响应(DR)计划，在压力较大的条件下远程关闭空调机组。虽然在许多情况下是有效的，但它们容易受到重大弱点的影响。首先，DTR通常是启发式的或试验性的，通常不适用于封闭形式的数学计算。此外，基于空调的灾难恢复通常是根据公用事业公司和用户之间的合同协议实施的，并且确实考虑了用户的幸福感(即，用户将因空调关闭而体验到的室内温度)。在严重的热浪事件下，这可能会导致负面的健康影响，特别是对婴儿和老年人。这项提议的目标是提高这两种工具的有效性。建议的解决方案模拟了温度过高对部件可用发电/输电能力的影响，以及对由于过载或在恶劣条件下运行而预期的部件寿命缩短的影响。通过开发住宅的热模型，住宅的室内温度被纳入到DR调度中，该模型可以基于内部和外部增益来确定室内温度。这就产生了一个多目标设计问题，其目标是在优化资产寿命和用户舒适度的同时优化成本。为了解决模型中固有的不确定性，采用了稳健优化方法。为了确保优化问题的易处理性和所提出解的可伸缩性，将使用标准的重构技术将非线性公式转换为凸的混合整数二次约束规划问题。此外，为了使用户能够了解情况，将建立基于非侵入性监测的算法，以使用智能电表提供的综合测量来检测空调设备的人员占用和状态。

Extreme temperatures can push various power grid components to their operational limits. The available capacity of most generation resources and power system components becomes negatively affected as the temperature increases beyond certain thresholds. Not surprisingly, this temperature-induced reduction in available power generation and transmission capacities generally coincides with increased electricity demand on the system, mostly attributed to the increased utilization of air-conditioning (A/C) systems. Ignoring the effects of temperature on various grid assets could lead to overloading these assets, resulting in reduced lifetime and premature component failure. It is therefore crucial to incorporate the effects of ambient temperature into power grid operation in order to prevent stress on components and avoid blackouts that could result from failure to meet demand during periods of extreme temperature. This issue is becoming more important since climate models project an increase in the duration and frequency of heat waves. Loss of power during extreme temperature conditions is not merely an inconvenience, as it may also impact the availability of other critical infrastructures such as water sanitation plants, transportation systems, and hospitals and other urgent care units. In this project, the researchers will pursue a possible solution that involves design of a methodology for proactive dispatch of the energy resources in a distribution system exposed to extreme ambient temperatures. Electric utilities have traditionally addressed the issue at hand through two means: defining dynamic thermal ratings (DTR) for various components to adjust their available capacity based on ambient temperature and, more recently, offering incentivized demand response (DR) programs for remotely shutting down A/C units under stressed conditions. Although effective in many instances, are vulnerable to significant weaknesses. First, DTR are often assigned heuristically or experimentally, and are not usually amenable to closed form mathematical calculation. Also, A/C-based DR is usually implemented based on the contractual agreements between the utility and the users, and does incorporate users' well-being (i.e., the indoor temperature users will experience due to A/C shutdown). This, under severe heat wave events, can potentially lead to negative health impacts especially on infants and the elderly. The goal of this proposal is to improve the effectiveness of both these tools. The proposed solution models the effects of excess temperatures on available generation/transmission capacity of components, as well as on expected reduction in component lifespan due to overloading or operating under harsh conditions. Indoor temperatures at residential homes are incorporated into the DR dispatch by developing thermal models for houses, which can determine the indoor temperature based on internal and external gains. This creates a multi-objective design problem in which the aim is to optimize cost in conjunction with asset lifetime and user comfort. To address the inherent uncertainties in the model, a robust optimization approach is adopted. To ensure tractability of the optimization problem and the scalability of the proposed solution, standard restructuring techniques will be used to transform the nonlinear formulation into a convex, mixed-integer quadratically constrained programming problem. Furthermore, to enable awareness on user conditions, algorithms will be built based on non-invasive monitoring to detect human occupancy and status of A/C units using the aggregate measurements available from smart meters.