UC1 Buildings efficiency
Contents
General description of use case
With 40% of total energy consumption, buildings are significant consumers of energy, causing 36% of the total CO2 emissions. Thereby, the building sector has enormous untapped potential for reducing CO2 emissions by optimising construction and management. The EU has issued a Directive on the Energy Performance of Buildings (EPBD) as the main EU policy instrument to improve in this regard (Directive 2002/91/EC). The directive involves a framework for assessing the energy performance of buildings through Energy Performance Certificates (EPCs) that need to include reference values, such as current legal standards. Hence, the relationship between buildings and energy impacts a broad spectrum from strategic to operational concerns, including energy efficiency investment decisions, buildings energy performance management, energy efficiency policies, smart buildings, and energy disaggregation. This makes data on energy efficiency in buildings crucial for discussion and decision-making in many practical contexts by households, academics, or policymakers.
Moreover, a wide variety of information is needed, from the existing building stock to ventilation and air conditioning solutions, socio-demographic information, cultural perception of thermal comfort to climatic and weather information. To some extent, this data is available; however, databases are unorganised and not interlinked. A comprehensive collection of European buildings and urban stock data has been identified by the European Buildings Stock Observatory, serving as a starting point for the use case.
A specific part of energy consumption in residential buildings is due to the use of household appliances. Time series of corresponding demand profiles on this scale are typically not open but constitute business assets of utilities. A wealth of information on such data is produced due to the massive rollout of smart meters across Europe. Access to this data raises essential concerns regarding privacy issues. Thus, legal restrictions, as well as commercial interests, have to be accounted for. Finally, analysing this information by, e.g., geolocation or social strata is critical to mitigating energy poverty, making the energy transition inclusive for all.
The use cases in EERAdata are selected and designed to (a) cover essential aspects of data-driven low-carbon energy research, (b) test and refine the FAIR/O ecosystem with a focus on interdisciplinary and complex data, (c) take up expert knowledge from different energy fields, so that taxonomies and metadata follow the mental model of experts for their fields, allowing mapping of conceptualisations developed in different domains of low carbon energy research, and (d) suggest FAIR/O data standards in the field of use cases and to produce methodological blue-prints for approaching other use cases in the future.
Technical challenges
Users of energy data include researchers, industry representatives, policymakers, funding and publishing agencies, and the general public. These users have different expectations and requirements for the format, type, granularity, and scope of data. With the changing structure of the energy system, more automated decision and control systems (machines) are in place to support humans in managing the energy infrastructure. Hence, it becomes necessary to design, process, and share energy data in a machine-actionable format. For the buildings data, the need to engage with many user groups with varying requirements, the lack of standards for data and metadata, and the lack of common community platforms for information pose significant challenges.
Data standards bring a unified form for (meta)data, enhance their FAIRness, contribute to sharing data between different user types, enable transparency, and improve the understanding through common meanings, consistency, and quality of the data.
Societal challenges
Buildings are responsible for 36% of the total CO2 emissions, and 40% of total energy consumption. On the other hand, this points to a significant potential for decreasing CO2 emissions, for instance, by increasing energy savings in buildings. Likewise, the relationship between buildings and energy is essential from a variety of aspects relevant to emissions mitigation and climate change adaptation. These include energy efficiency investment decisions, buildings energy performance management, energy efficiency policies, smart buildings, and energy disaggregation. Hence, data on energy efficiency in buildings is crucial for the discussion and decision-making in many practical contexts by households, researchers, or policymakers.
This data originates from a multitude of sources and does not conform to common standards, hindering Interoperability and Reusability. Moreover, there are significant concerns regarding privacy issues, legal restrictions, and commercial interests.
Conclusions
The contemporary paradigm of the energy system poses challenges regarding the implementation of the FAIR principles. The foremost challenge is associated with the many different types of stakeholders and users, implying a wide variety of data needs. Researchers might be more interested in past data, whereas actors of the electricity market would require real-time data at high resolutions. Clearly, policy-makers need to track, collect and process data for their decisions and policy-making. Households would base their behaviours on energy prices, whereas funding agencies would follow trends and metrics in the energy field.
Given the different needs of different user types, another challenge for implementing FAIR principles is adapting to the energy system –and the energy data – to the ‘new’ players of the system. These automated decision and control systems (machines) support humans in managing the energy infrastructure. These emerging new actors of the energy system make it even more urgent to design, process, and share energy data in a machine-actionable format. That is, in a format such that machines can (be programmed to) find, access, and process data without or with very little human intervention. Hence, the new paradigm of the energy system calls for FAIRness and machine-actionability.
Although there is not a single common way to achieve this, the first operational step would be “Define”, i.e., identifying the current status and needs of energy data in terms of compliance with the FAIR principles. That is, FAIR implementation requires first evaluating the current status of the FAIRness of data and metadata in the energy domain. The initial observations concerning the FAIR status of the data and metadata in the Buildings
Efficiency domain highlights several challenges. A first review of the buildings efficiency data reveals that part of the data needed by users, such as usage rates of electricity in buildings is captured by utility companies, mainly for billing purposes. Different users might utilize this data, for instance by analysing load patterns and developing suggestions for energy conservation. However, this data is usually proprietary for the utility companies, and does not satisfy Findability, Accessibility, Reusability, or Openness criteria. Likewise, the vast amount of data generated by energy-related projects focus on buildings, households, their habits and preferences through field studies. Such data is also kept in the project repositories, and fail to be Findable, Accessible, or Open. Moreover, such data is usually not Accessible beyond the projects’ lifetime. A more practical FAIRness issue concerns the smart meters, appliances or sensors in buildings. In many cases, the data structures of these smart systems from different manufacturers are not compatible.
Owing to the privacy concerns, the data is not Open and Accessible. Apart from the practical concerns, a detailed assessment of the buildings energy data reveals that, in terms of Findability, there are severe limitations concerning buildings efficiency data as most data is kept locally. Regarding Accessibility, there are severe issues concerning the maintenance of the data repositories over time, especially beyond the associated project lifetimes. Interoperability is also significantly affected from different, and in some cases unexploitable data structures (e.g., qualitative data such as text or audio format interview transcripts) that pose barriers for the Interoperability of buildings efficiency data. The Reusability of the data is impacted by privacy concerns, limitations exercised by data sources, missing data, multiplicity and scattered nature of data sources (households, industries, utility companies, municipalities), as well as how well the data characteristics are defined and reported. The Openness of the data also comes in as a significant drawback, mainly due to restrictions arising from commercial and security concerns.
The second phase in the FAIRification roadmap is the “Implement” step. As with defining data, metadata, and their FAIRness assessments, a single recipe cannot be defined for the implementing FAIR. The experience from the EERAdata project reveals that, the energy research community needs to define their own way of implementing the FAIR data principles. The efforts for the energy and buildings efficiency data started with discussing and developing metadata standards for FAIR and open data in the low carbon energy research community. These efforts focused on the link between metadata and how humans explore data. In result, gaps and needs that hinder their standardised implementation were identified. In result, the challenges identified were costs associated with handling data, issues with ensuring the Accessibility of data in the long term, difficulty of merging qualitative and quantitative data, lack of sufficient technical capacity to share qualitative data, and lack of awareness concerning the value of FAIR data. These challenges suggest that the community needs first to reach to a consensus on the value and costs of FAIR data. Assessing the value of FAIR data amounts to identifying the economic and social benefits that can be generated through the FAIRification of data. In contrast, the costs pertain to the invested resources (time and human). Hence, for the “Implement” phase, the suggestions can be listed as focusing on community or domain-wide standardisation, utilisation of larger platforms for hosting data instead of smaller individual projects platforms, integrating or establishing common repositories for similar databases, and utilising repositories for metadata augmentation.
The third phase in the EERAdata FAIRification roadmap is “Embed and sustain” where business models, licensing, metrics, and certification are exploited to highlight the value of FAIRification of data. The value of FAIR and Open data can be exemplified by the added value that can be generated by sharing the data, cost savings due to timely sharing of the data, the possibility to avoid the costs associated with reproducing or replacing the shared data, the sales price of the shared data in the market, and prices from comparable data. On the other hand, the identified challenges are: data silos hindering interoperability within and among businesses, the negative impact on Findability and Reusability caused by not storing data and metadata together, use of proprietary data formats and proprietary software resulting in data governance lock-ins, lack of universal access to sensor-based real-time data, lack of standardization and standards hidden behind paywalls, as well as the lack of energy data markets. At this point, a critical recommendation is to accomplish an apriori awareness and consensus regarding both the value that can be generated out of FAIR data, and the costs of FAIR data. This is a challenging task to achieve, however, it is key to mining the value and handling the costs of FAIRification.
Clearly, the value of data is a function of its attributes (e.g., FAIR status) as well as its design, how it is generated, processed, and shared. Hence, the FAIR and Open principles, including and implying machine-actionability, need to be incorporated into the design process, whether the data is generated through a research project, through a commercial practice, or by policy makers. This, in turn, requires the involvement of the potential users and their needs into all steps of the data lifetime.
In addition to the technical aspects of FAIRification, there is also the urgent need to identify and address the challenges associated with privacy concerns, ethics, and data protection requirements. That is, the energy research community also needs to establish procedures and standards for the processes involving the design, generation, processing, and sharing of data so that these processes produce FAIR data while respecting the ethical, privacy, and data protection concerns. One significant counterpart of these processes involve trust and the democratisation of knowledge.
Recommendations
Findability - Incorporate the FAIR and Open principles into the design process. - Utilize of larger platforms for hosting data instead of smaller individual projects platforms. - Promote community platforms for hosting domain-specific data - User-centric design of data and data practices/processes
Accesibility - Identify and address the challenges associated with the privacy concerns, ethics, and data protection requirements. - Establish procedures and standards pertaining to the processes involving the design, generation, processing, and sharing of data so that these processes produce FAIR data while respecting the ethical, privacy, and data protection concerns. - Design means of keeping the data Accessible after project lifetimes
Interoperability - Focus on community or domain-wide standardisation. - Integrate or establishing common repositories for similar databases, and utilising repositories for metadata augmentation. - Develop means of Interoperability for qualitative data
Reusability - User-centric design of data and data practices/processes - Promote trust within the community and the democratisation of knowledge. - Accomplish an apriori awareness and consensus regarding both the value that can be generated out of FAIR data, and the costs of FAIR data. - Develop means of Reusing qualitative data
