Optimised Demand Side Management for Flexible Operation of Smart Grids

为了实现我们目前的净零碳目标，电力系统必须容纳大量间歇性可再生能源和新负荷（例如，电加热和运输）。然而，升级能源系统的容量以适应所有上述技术将是昂贵的。在实践中，仅在峰值条件或极端条件下才需要能源系统的全部容量，例如，在冬季寒冷的日子里，当供暖需求最高或发生意外事故时，系统以降低的容量运行。这是因为，由于能源需求相对可预测和稳定，能源系统在历史上一直以被动的方式（适应和忘记）运行。由于新的（有时是间歇性的）技术连接到电网，情况不再是这样，我们需要更积极的方法来管理我们的能源系统，例如通过部署需求侧管理。需求侧管理，即客户通过使用设备（例如，存储）或通过改变它们的行为（例如，在晚上洗衣服）提供了有吸引力的选项来管理网络压力并减少对昂贵的网络容量的需求。然而，需求侧响应的使用可以在客户偿还其能量时对网络引入非预期的影响，例如，这些回报效应必须在部署需求侧响应时进行适当的建模和评估。本研究项目的目的是提供更准确的方法来捕捉回报效应，并提出部署需求侧响应的新方法。该项目的具体目标包括：-利用与电网分析工具兼容的模型（例如，ZIP模型）。探索适合需求侧建模和电力网络建模的不同优化技术，例如，多阶段优化，其中上层是线性的。生成相关的优化模型，并将其扩展到考虑实际的多周期（滚动范围）条件，其中随着收到更好的预测，需求侧响应部署不断重新优化。将该方法扩展到不同的电力网络级别，特别是网络不平衡往往是关键的低电压级别。该项目的产出旨在促进需求侧响应的部署（例如，由分销网络运营商、聚合商等）并以更具成本效益的方式协助我们达致能源减碳目标。

In order to meet our current net-zero carbon targets, the electricity system must accommodate large volumes of intermittent renewable energy sources and new loads (e.g., electrified heating and transports). However, upgrading the capacity of the energy system to accommodate all the aforementioned technologies would be costly.In practice, the full capacity of the energy system is only required during peak conditions,or extreme conditions, e.g., cold days in winter when heating demand is the highest or when a contingency occurs and the system operated with reduced capacity. This is because, the energy system has historically operated in a passive manner (fit-and-forget) because energy demand has been relatively predictable and stable. As this is no longer the case due to the new, sometimes intermittent, technologies being connected to the electricity grid, we need more active approaches to manage our energy systems, such as by deploying demand side management.Demand side management, where customers actively change their energy through the use of devices (e.g., storage) or by changing their behavior (e.g., washing clothes at night), is known to provide attractive options to manage network stress and reduce needs for costly network capacity. However, the use of demand side response can introduce unintended effects on the network as customers payback their energy, e.g., increasing stress during the periods when they charge their storage, do laundry, etc. These payback effects must be properly modelled and assessed when deploying demand side response.The aim of this research project is to provide more accurate means to capture the payback effects, and propose new methods to deploy demand side response. The specific objectives of this project include:- Capturing the characteristics of the different resources used to provide demand side response using models that are compatible with electricity network analysis tools (e.g., ZIP models).- Exploring different optimisation techniques that are suitable for both demand side modelling and electricity network modelling, e.g., multi-stage optimisations where the upper levels are linear.- Producing a relevant optimization model and extending it to consider realistic multi period (rolling horizon) conditions where demand side response deployment is constantly re-optimised as better forecasts are received.- Extend the approach to different electricity network levels, especially to the low voltage level where network imbalances tend to be critical.The outputs of the project are meant to facilitate deployment of demand side response (e.g., by distribution network operators, aggregators, etc.) and facilitate meeting our energy decarbonisation targets in a more cost-effective manner.