Data-Smart Building Case Studies

建筑行业约占最终能源使用总量的40%，并蕴藏着以具有成本效益的方式节约能源和减少二氧化碳排放的巨大潜力。其中许多具有成本效益的机会不需要大量的资本支出，而是依赖于合理的决策和熟练地实施建筑物维护和运行控制战略。事实上，许多建筑从一开始就委托得很差，以至于它们从未像设计的那样运行。即使对于使用良好的建筑，系统的性能也可能随着时间的推移而下降，这可能会被忽视，从而导致性能低下。Katipamula和Brambley(2005)强调了这种机会的程度，他们声称“维护不善、退化和控制不当的设备浪费的能源估计占商业建筑所用能源的15%到30%。”这些问题的重要性现在得到了广泛的承认。例如，在关于负担得起的供暖和制冷的特派团创新挑战7中，预见了一项关于预测性维护和控制优化(PMO)的具体任务。在这项任务中取得进展的一个关键先决条件是获得高质量的背景数据和计量数据。在英国，智能电表研究门户项目寻求通过将上下文链接到智能电表数据来创建和公开此类资源。丰富的数据集正变得越来越多，但问题仍然是如何最大限度地从这些数据中提取洞察力和可操作的信息。该项目计划的工作旨在探索如何将这些数据集有益地用于解决项目管理办公室的问题；此外，它还将寻求探索更综合的建筑物级集成的好处。主要前提是可以处理上下文信息和高度精细的数据，以识别性能降级并为预测性维护决策提供信息。因此，寻求通过更智能的控制策略来调整和优化运营是有意义的。除了技术挑战之外，为了有效地采用这种以数据为中心的新方法，还需要确定各种利益攸关方的价值主张，并确定潜在可行的业务模式、监管边界和采购机制。在世界各地，出现了与数据智能建筑的好处和实际应用途径有关的良好实践范例和证据。这项工作的范围可能很广，不能在一个单一的项目中解决。最近制定的国际能源署EBC附件81，数据驱动的智能建筑寻求汇集世界各地的资源，创造一个临界量的研究人员，然后可以应对这样的挑战。该项目将支持联合王国在附件D分任务D案例研究和商业模式方面的参与和领导。将收集的案例研究将考虑从测量(收集数据)到建筑管理(可操作的知识和建筑控制)过程中的逻辑步骤。这些方面将包括最新的研究进展，包括：(I)数据收集和数据建模；(Ii)数据驱动的建模；(Iii)获取专业知识，但也使用人工智能来创建数据驱动的应用程序和服务；(Iv)使用和采用新的服务和商业模式。这种方法遵循了一句谚语：“你不能管理你不衡量的东西”。为了确保证据的相关性，这项工作将涵盖一系列具有代表性的建筑类型、气候和居住者应用。调查结果将传达给利益相关者。

The building sector accounts for about 40% of total final energy use and harbours the enormous potential to save energy and reduce CO2-emissions in a cost-effective way. Many of these cost-effective opportunities do not require significant capital outlay, relying instead on sound decision making and proficient implementation of building maintenance and operational control strategies. Indeed, many buildings are poorly commissioned from inception such that they never operate as designed. Even for well-commissioned buildings, the performance of systems might degrade over time, and this can go unnoticed resulting in poor performance. The extent of the opportunity is highlighted by Katipamula and Brambley (2005) who claim that "poorly maintained, degraded, and improperly controlled equipment wastes an estimated 15% to 30% of the energy used in commercial buildings." The importance of these issues in now widely acknowledged. For example, within the Mission Innovation Challenge 7 on Affordable Heating and Cooling a specific task on Predictive Maintenance and control Optimization (PMO) is foreseen. A key pre-requisite to achieve progress in such a task is access to high-quality contextual and metered data. In the UK, the Smart Meter Research Portal project seeks to create and make publicly available such a resource by linking contextual to smart meter data. Rich data sets are becoming increasingly available, but the question remains on how to maximise the insights and actionable information that can be extracted from these data.The work planned for this project seeks to explore the ways in which such datasets can be beneficially used to address the issue of PMO; in addition, it will seek to explore the benefits of more integrated building-level integration. The main premise is that contextual information together with highly granular data can be processed to identify performance degradations and inform predictive maintenance decisions. It then makes sense to seek to tune and optimise operation by more intelligent control strategies. Beyond technical challenges, for effective adoption of such new data-centric approaches, the value proposition needs to be identified for various stakeholders, and the identification of potentially viable business models, regulatory boundaries and procurement mechanisms. Around the world, good practice examples and evidence are appearing related to the benefits and practical application pathways towards Data-Smart Buildings.The scope of this work is potentially vast and cannot be addressed within a single project. The recently established IEA EBC Annex 81, Data-Driven Smart Buildings seeks to pool resources from around the world to create a critical mass of researchers then can address such a challenge. This project will support the UK's involvement and leadership on Subtask D of the Annex, on Case Studies and business models. The case studies to be collected will consider the logical steps on the journey from measurement (gathering data) to building management (actionable knowledge and building control). These aspects are to include latest research developments that span from: (i) data collection and data-modelling; (ii) data-driven modelling; (iii) capturing expert knowledge but also using Artificial Intelligence, to create data-driven applications and services, to; (iv) the utilisation and adoption of new services and business models. This approach follows the adage that 'you can't manage what you don't measure'. To ensure that the evidence is relevant, the work will cover a range of representative building typologies, climates and occupant applications. The findings will be communicated to stakeholders.