Technology Transformation to Support Flexible and Resilient Local Energy Systems

能源系统的供应面正在发生深刻的变化。英国已将电力系统的碳排放量从2012年的1.5亿吨以上减半至2018年的70吨以下，中国正在每年增加约2000万千瓦的风力发电能力，并已将深圳市所有公交车更换为电动公交车。很明显，还有更多工作要做：剩余的电力脱碳，其他部门的电气化，以及寻找替代的零碳燃料。城市传统上一直是从腹地带来的能源的巨大消耗者，但能源网络的负荷增长是不可避免的，因为更多的服务，特别是交通，通过电气化来脱碳，以及通过重新开发更高的建筑来增加建筑密度。对此的传统反应是增加更多的厂房和设备，这被认为是一种低效的做法。一个有趣的趋势是当地能源系统(LES)和多能源微电网(MEMG)的出现。LES和MEMG是一种手段，可以提高当地能源资源的自我消耗；利用能源服务的灵活性来源；利用地方服务进行地方和国家控制，并转向确保复原力的智能方法。英国最近的停电(9/8/19)突显出交通系统和其他城市基础设施特别脆弱。重新思考如何在城市环境中提供弹性可以极大地减少这种脆弱性。尽管中国和英国城市的历史和地理不同，但我们发现了共同的挑战和一套互补的研究专业知识。这个项目汇集了来自英国和中国的电力电子、优化、控制和故障管理方面的专家。现有的能源网络，特别是电力网络，是在假设电力从偏远的发电站进入城市并由城市内部的网络分配的情况下设计的。这导致了从变电站向外延伸的一组放射状线路。这种结构无法支持围绕本地发电和存储资源形成灵活的微电网。我们建议使用电力电子设备来重构传统网络，这些设备可以在以前未连接的网络点之间提供受控的功率流。这开启了城市地区电力流动的动态重组，以允许更大的本地能源自我消耗，例如将太阳能住宅物业转移到电动汽车的工作场所充电。它还允许在停电的情况下形成岛屿，以保持基本服务的运行，同时搁置非基本服务。我们还研究了硬件和控制问题。原则上已经讨论了电力电子路由的硬件，但它太大了，效率不够高，不能在城市环境中使用。我们将致力于新形式的模块化功率转换器，以提高效率，减少物理体积，并提供对组件故障的弹性。能源网络的控制系统是集中的：它们从大片地区收集数据，做出决策，然后发布命令。微电网的概念将这一点改变为一种分散化的方法。分散化的一个关键好处是可以随时获得有关能源消耗灵活性的信息，例如哪些电动汽车可以推迟充电或可能提供电力以帮助停电。地方控制还提供了将建筑物的供热/制冷、运输能源系统和电力系统作为一个综合整体运行的机会。这可以导致更好地整合可再生能源，从而深度脱碳，但需要在管理当地能源资源和需求的不确定性方面向前迈出重大一步。我们将利用随机优化和机器学习技术来解决这个问题，并为智能城市能源系统设计一个控制和运行框架。

Deep changes are happening in the supply side of energy systems. The UK has halved carbon emissions from electricity system from over 150 million tonnes in 2012 to under 70 in 2018 and China is adding about 20 GW of wind generation capacity per year and has replaced all buses in the city of Shenzhen with electric busses. Very clearly there is much more to do: the remaining decarbonisation of electricity, the electrification of other sectors and sourcing alternative, zero-carbon fuels.Cities have traditionally been huge consumers of energy brought in from their hinterland and yet load growth in energy networks is inevitable as more services, notably transport, are decarbonised through electrification and building density increases through re-development of with taller buildings. The traditional response to this, adding more plant and equipment, is recognised as being an inefficient. An interesting trend is the emergence of Local Energy Systems (LES) and Multi-Energy Micro-Grids (MEMG). LES and MEMG are a means for raising self-consumption of local energy resources; tapping into sources of flexibility in how the services derived from energy; using local services for both local and national control and moving to a smart ways of ensuring resilience. The recent power outage in the UK (9/8/19) highlighted that transport systems and other urban infrastructure are particularly vulnerable. A re-imagining of how resilience is provided in the urban setting could hugely reduce that vulnerability.Despite the differences between the histories and geographies of cities in China and the UK, we find common challenges and a complementary set of research expertise. This project brings together experts in power electronics, optimisation, control and fault-management from UK and China.Existing energy networks, especially electricity networks, were designed assuming power enters a city from remote power stations and the network inside the city distributes this. This led to a radial set of lines spreading out from substations. This structure is unable to support the formation of flexible microgrids around local generation and storage resources. We propose to re-structure the legacy networks using power electronics devices that give controlled power flows between previously unconnected networks points. This opens up dynamically restructuring the power flow in urban areas to allow greater local self-consumption of energy, for instance moving solar power residential properties to work-place charging of electric vehicles. It also allows islands to be formed in reaction to power cuts that keep essential services running while placing non-essential services on hold. We also look at hardware and control issues. The hardware for electronic routing of power has been discussed in principle but it is too large and not efficient enough to be used in urban settings. We will work on new forms of modular power converter that raise efficiency, reduce physical volume and provide resilience to component failures. Control systems for energy networks are centralised: they gather data from across large areas, make decisions and then issue commands. The microgrid concept changes this to a decentralised approach. A key benefit of decentralisation is the ready access to information about flexibility in energy consumption, e.g which electrical vehicles could delay charging or might supply power to aid with a power cut. Local control also gives opportunities to run the heat/cooling of buildings, the transport energy system and the electricity system as an integrated whole. This can lead better integration of renewable energy and therefore deep decarbonisation but requires a major step forward in managing uncertainty over the local energy resources and demands. We will bring the techniques of stochastic optimisation and machine learning to bear on this problem and devise a control and operations framework for smart urban energy systems.